Report of the Motherisk
Hair Analysis
Independent Review

The Honourable Susan E. Lang
Independent Reviewer

© Ontario Ministry of the Attorney General 2015

Published by the Ontario Ministry of the Attorney General

ISBN 978-1-4606-7157-3 (PDF)
ISBN 978-1-4606-7156-6 (HTML)

www.m-hair.ca
www.attorneygeneral.jus.gov.on.ca

Also available in EPUB format

Disponible en français

Letter of transmittal - header

December 15, 2015

The Honourable Madeleine Meilleur
Attorney General of Ontario
Ministry of the Attorney General
McMurtry-Scott Building
720 Bay Street, 11th Floor
Toronto, ON M7A 2S9

Re: Motherisk Hair Analysis Independent Review

Dear Minister:

With this letter, I am delivering the Report of the Motherisk Hair Analysis Independent Review.

I have concluded that the hair-strand drug and alcohol testing used by the Motherisk Drug Testing Laboratory between 2005 and 2015 was inadequate and unreliable for use in child protection and criminal proceedings and that the Laboratory did not meet internationally recognized forensic standards. The use of the Laboratory’s hair-testing evidence in child protection and criminal proceedings has serious implications for the fairness of those proceedings and warrants an additional review.

It has been a privilege to serve as the Independent Reviewer. I thank you for the support provided by your Ministry.

Yours truly,

Susan Lang signature
Susan Lang
Independent Reviewer

Letter of transmittal - footer

Table of Contents

  1. Abbreviations and Acronyms
  2. Glossary
  3. CHAPTER 1
    Executive Summary
    1. Introduction
    2. The Laboratory and the Independent Review
    3. Use for Forensic Purposes
    4. Adequacy and Reliability
      1. Analytical Practices: 2005–10
        1. ELISA Tests
        2. Absence of Standard Operating Procedures and Contemporaneous Documentation
        3. Additional Flaws
      2. Analytical Practices: 2010–15
      3. Analytical Practices: Alcohol Markers
      4. 2005–15: Interpretation and Communication Practices
        1. Misuse of Concentration Ranges
        2. Poor Communication to Customers
      5. Proficiency Testing and Accreditation
        1. Proficiency Testing
        2. Accreditation
      6. At No Time Did MDTL Meet Forensic Standards
    5. Oversight Failures at the Hospital for Sick Children
    6. Child Protection and Criminal Proceedings
    7. Summary of Recommendations
    8. Conclusion
  4. CHAPTER 2
    The Independent Review: Origins, Mandate, and Process
    1. 1. Introduction
    2. 2. Origins of the Review: The Broomfield Case and Aftermath
    3. 3. Original and Expanded Mandate
    4. 4. Setting Up the Review
      1. 4.1 Review Counsel
      2. 4.2 Scientific Experts
      3. 4.3 Administrative and Editorial Support
    5. 5. Process of the Review
      1. 5.1 Primary Document Review
      2. 5.2 Interviews with Current and Former MDTL Staff
      3. 5.3 Consultations with Organizations
      4. 5.4 Submissions
      5. 5.5 Examination of Court Decisions
      6. 5.6 Legal Experts and Roundtables
    6. 6. Nature of Findings
  5. CHAPTER 3
    Introduction to Hair Analysis
    1. 1. Introduction
    2. 2. Hair Testing as a Method of Drug or Alcohol Testing
      1. 2.1 Distinguishing between Use and Exposure for Drugs of Abuse
      2. 2.2 Bulk Analysis vs. Segmental Analysis for Drugs of Abuse
      3. 2.3 Testing for Alcohol Markers in Hair
    3. 3. The Hair-Testing Process
      1. 3.1 Sample Collection
      2. 3.2 Sample Preparation and Analysis
        1. 3.2.1 Sample Preparation
          1. 3.2.1.1 Segmentation
          2. 3.2.1.2 Washing
          3. 3.2.1.3 Homogenization
        2. 3.2.2 Sample Pre-treatment and Extraction
        3. 3.2.3 Testing for Drugs of Abuse
          1. 3.2.3.1 Immunoassay-Based Screening Tests
          2. 3.2.3.2 Confirmation Tests
          3. 3.2.3.3 A Note on Terminology
        4. 3.2.4 Testing for Alcohol Markers
      3. 3.3 Analytical Concepts
        1. 3.3.1 Calibration
        2. 3.3.2 Quality Controls
        3. 3.3.3 Internal Standards
        4. 3.3.4 Cut-offs
      4. 3.4 Quality Management System
    4. 4. Interpretation of Hair Test Results
      1. 4.1 Hair Colour Bias
      2. 4.2 Effect of Hair Products, Cosmetic Treatments, and Hair Damage
      3. 4.3 Interpreting Timing of Use or Exposure
        1. 4.3.1 Variation in Hair Growth Rates
        2. 4.3.2 Banding
        3. 4.3.3 Sample Collection
        4. 4.3.4 Potential Residual Concentrations
      4. 4.4 Potential for External Contamination for Drugs of Abuse
        1. 4.4.1 Interpreting Drug Metabolites
        2. 4.4.2 Analyzing Washings
      5. 4.5 Dose–Response Relationship and Inter-subject Variation
      6. 4.6 Segmental Analysis and Serial Testing for Drugs of Abuse
      7. 4.7 Infant- and Child-Specific Considerations for Drugs of Abuse
      8. 4.8 Interpreting Body Hair
      9. 4.9 Interpreting Alcohol Markers
      10. 4.10 Understanding the Limitations of Hair Analysis
  6. CHAPTER 4
    Structure and Evolution of MDTL
    1. 1. Introduction
    2. 2. Situating MDTL within the Hospital for Sick Children
      1. 2.1 The Motherisk Program
      2. 2.2 The Research Institute
      3. 2.3 The Department of Paediatric Laboratory Medicine
      4. 2.4 The Division of Clinical Pharmacology and Toxicology
    3. 3. Structure of MDTL
      1. 3.1 Laboratory Director
      2. 3.2 Laboratory Manager
      3. 3.3 Laboratory Supervisor
      4. 3.4 Other Staff
        1. 3.4.1 Laboratory Technicians and Laboratory Technologists
        2. 3.4.2 Research Associates and Research Technologists
        3. 3.4.3 Quality Manager
      5. 3.5 Graduate Students and Fellows
    4. 4. Understanding Forensic vs. Clinical vs. Research Purposes
    5. 5. Evolution of MDTL’s Hair-Testing Work
  7. CHAPTER 5
    Adequacy and Reliability of Analytical Methods Used by MDTL
    1. 1. Introduction
    2. 2. Understanding Internationally Recognized Forensic Standards
    3. 3. Drugs of Abuse: January 2005–August 2010
      1. 3.1 Improper Use of ELISA Results
        1. 3.1.1 ELISA vs. GC-MS or LC-MS/MS Comparisons
      2. 3.2 Absence of Standard Operating Procedures and Contemporaneous Documentation
        1. 3.2.1 Sample Suitability
        2. 3.2.2 Washing
        3. 3.2.3 Extraction Method
        4. 3.2.4 Calibrators
        5. 3.2.5 Quality Controls
        6. 3.2.6 Criteria for Acceptability
        7. 3.2.7 Cut-offs
      3. 3.3 Absence of Oversight
      4. 3.4 Failure to Wash Samples Routinely
      5. 3.5 Inadequate Chain-of-Custody Procedures
      6. 3.6 Inadequate Record Keeping
    4. 4. Drugs of Abuse: September 2010–April 2015
      1. 4.1 Improvements in MDTL’s Analytical Practices
        1. 4.1.1 GC-MS Confirmation and Elimination of Quantitative ELISA Results
        2. 4.1.2 Written Standard Operating Procedures and Contemporaneous Documentation
        3. 4.1.3 Improved Oversight
        4. 4.1.4 Routine Washing of Adult Samples
        5. 4.1.5 Complete Case Files
      2. 4.2 Inadequacies in MDTL’s Analytical Methods
        1. 4.2.1 Deficiencies in GC-MS Procedure
          1. 4.2.1.1 Identification Method
          2. 4.2.1.2 Quantification Method
        2. 4.2.2 Continued Reliance on Unconfirmed ELISA Results
        3. 4.2.3 Routine Failure to Wash Neonatal and Child Samples
        4. 4.2.4 Inadequate Chain-of-Custody Procedures
    5. 5. Alcohol Markers: 2005–15
      1. 5.1 Evolution in the Science of Alcohol Marker Testing in Hair
      2. 5.2 Evolution in MDTL’s Analytical Methods
      3. 5.3 Adequacy and Reliability of MDTL’s Analytical Methods
        1. 5.3.1 Improvements over Analytical Methods for Drugs of Abuse
        2. 5.3.2 Deficiencies in Analytical Methods for FAEEs
        3. 5.3.3 Changes in the Science
  8. CHAPTER 6
    Adequacy and Reliability of Interpretations and Opinions Given by MDTL
    1. 1. Introduction
    2. 2. MDTL’s Use of Concentration Ranges
      1. 2.1 What Is a Concentration Range?
      2. 2.2 Why MDTL Created Concentration Ranges
      3. 2.3 Concentration Ranges Should Not Be Used Forensically
        1. 2.3.1 They Were Built on Flawed Data
        2. 2.3.2 They Did Not Account for Inter-subject Variation
      4. 2.4 The Laboratory Was Aware of the Pitfalls of Concentration Ranges
      5. 2.5 Despite Pitfalls, MDTL Used Concentration Ranges Inappropriately
      6. 2.6 Concentration Ranges in Use: September 2010–15
    3. 3. MDTL’s Communication of Its Test Results
      1. 3.1 Nobody Had the Expertise to Give Forensic Interpretations
      2. 3.2 MDTL’s Interpretations Not Appropriate
        1. 3.2.1 MDTL Did Not Provide an Interpretation in Each Case
        2. 3.2.2 MDTL Rarely Provided Written Interpretations
      3. 3.3 MDTL Did Not Communicate the Limitations of ELISA
      4. 3.4 MDTL Did Not Explain the Change in Testing Methodology
      5. 3.5 Issues Raised by Interpretations for Cocaine and Cannabis / THC
  9. CHAPTER 7
    Proficiency Testing and Accreditation
    1. 1. Introduction
    2. 2. MDTL and Proficiency Testing
      1. 2.1 Drugs of Abuse: Pre-2010
      2. 2.2 Drugs of Abuse: Post-2010
      3. 2.3 Alcohol Markers
    3. 3. MDTL and Accreditation
      1. 3.1 No Forensic Accreditation
      2. 3.2 Accreditation under the Ontario Laboratory Accreditation Program
        1. 3.2.1 Licensing and Accreditation Framework
        2. 3.2.2 Accreditation of MDTL under the Ontario Laboratory Accreditation Program
  10. CHAPTER 8
    What Went Wrong at SickKids
    1. 1. Introduction
    2. 2. Failure to Appreciate the Forensic Nature of MDTL’s Hair Tests
      1. 2.1 The Views of MDTL’s Leadership
      2. 2.2 The Views of the Hospital’s Leadership
    3. 3. Inadequate Training and Expertise at MDTL
    4. 4. Slow Transition from Research to Clinical Laboratory, but Never to Forensic Laboratory
      1. 4.1 Delay in Improvements to MDTL’s Analytical Methods and Operations
    5. 5. Lack of Transparency about MDTL’s Methods
    6. 6. No Meaningful Oversight of MDTL
  11. CHAPTER 9
    MDTL and Child Protection
    1. 1. Introduction
    2. 2. Overview of the Child and Family Services Act
      1. 2.1 The CFSA’s Purpose
      2. 2.2 The CFSA and Child Protection
    3. 3. Expert Evidence
    4. 4. MDTL Hair Tests and Child Protection Work
      1. 4.1 Use of MDTL Hair Test Results at Various Stages of Proceedings
      2. 4.2 The Introduction of MDTL’s Hair Test Results in Court
  12. CHAPTER 10
    MDTL and Criminal Proceedings
    1. 1. Introduction
    2. 2. Broomfield and MDTL Evidence
    3. 3. Opposing Expert Evidence on Broomfield
    4. 4. Conclusion
  13. CHAPTER 11
    Recommendations
    1. 1. Introduction
    2. 2. Recommendations for a Second Review – A Commissioner and a Review and Resource Centre
      1. 2.1 The Commissioner
      2. 2.2 The Review and Resource Centre
        1. 2.2.1 Counselling Services
        2. 2.2.2 Legal Resources
        3. 2.2.3 Alternative Dispute Resolution Services
      3. 2.3 Scope of the Second Review and Notice about Potentially Affected Persons
      4. 2.4 High-Priority Cases
      5. 2.5 Children Who Have Been Adopted
      6. 2.6 Resources
    3. 3. Additional Recommendations Regarding Criminal Cases
    4. 4. Financial Compensation
    5. 5. Recommendation for the Hospital for Sick Children
    6. 6. Recommendations for the Justice System
  14. Appendices
    1. APPENDIX 1Order in Council, November 26, 2014
    2. APPENDIX 2Order in Council, April 22, 2015
    3. APPENDIX 3Meetings with Organizations and Roundtables
    4. APPENDIX 4Submissions Received
    5. APPENDIX 5Reported Child Protection Cases Reviewed
    6. APPENDIX 6Further Reading on the Science of Hair Testing
    7. APPENDIX 7Hair-Testing Data, 2005–15, Provided by the Hospital for Sick Children
    8. APPENDIX 8Drug Concentration Ranges
      1. 8a Ranges of Concentration in Hair – Earliest Version
      2. 8b Ranges of Drug Concentration in Adult Hair – March 2006
      3. 8c Ranges of Drug Concentration in Infant Hair – March 2006
      4. 8d Ranges of Drug Concentration in Adult Hair – August 2011
      5. 8e Interpretation of Drug Detection in Child Hair – August 2011
    9. APPENDIX 9MDTL, The Use of Hair Testing to Establish Illicit Drug Use
    10. APPENDIX 10Monique Moller, Joey Gareri, Gideon Koren, “A Review of SubstanceAbuse Monitoring in a Social Services Context”
    11. APPENDIX 11MDTL Report Templates
      1. 11a Results Report Template – redacted and annotated by MDTL
      2. 11b Interpretation Report Templates – redacted – May 2007
      3. 11c Interpretation Report Template – Adult – July 2011
      4. 11d Interpretation Report Template – Child & Neonate – June 2014
    12. APPENDIX 12MDTL Interpretation Guidelines
      1. 12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009
      2. 12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009
    13. APPENDIX 13Alcohol Marker Concentration Ranges
      1. 13a Ranges of FAEE Concentration in Adult Hair – Earliest Version
      2. 13b Ranges of FAEE Concentration in Adult Scalp Hair – July 2007
      3. 13c Ranges of FAEE Concentration in Adult Scalp Hair – March 2010
      4. 13d Interpretation of Alcohol Hair Analysis, Adult Scalp Hair – March 2011
      5. 13e Interpretation of Alcohol Hair Analysis, Adult Scalp Hair – May 2011
    14. APPENDIX 14MDTL Letter to Clients, October 15, 2010
    15. APPENDIX 15Number of Adoptions from Individuals in Care, 2005–14
    16. APPENDIX 16Child Protection Proceedings Statistics, 2005-15
    17. APPENDIX 17Number of Individuals Tested by MDTL, 2005–15, Organized by Ontario Child Protection Agency
  15. Independent Reviewer and Review Staff

Abbreviations and Acronyms

6-AM

6-acetylmorphine

AAFS

American Academy of Forensic Sciences

ADR

alternative dispute resolution

ASCLD / LAB

American Society of Crime Laboratory Directors / Laboratory ­Accreditation Board

CALA

Canadian Association for Laboratory Accreditation

CFSA

Child and Family Services Act

DPLM

Department of Paediatric Laboratory Medicine

ELISA

enzyme-linked immunosorbent assay

EtG

ethyl glucoronide

FAEEs

fatty acid ethyl esters

GC-MS

gas chromatography–mass spectrometry

HS-SPME

headspace–solid-phase micro-extraction

IATDMCT

International Association of Therapeutic Drug Monitoring and ­Clinical Toxicology

IQMH

Institute for Quality Management in Healthcare

ISO

International Standards Organization

LC-MS/MS

liquid chromatography–tandem mass spectrometry

LLE

liquid-liquid extraction

LOD

limit of detection

LOQ

limit of quantification

xvi

LSD

lysergic acid diethylamide

MDA

methylenedioxyamphetamine, also known as 3,4-methylenedioxy-amphetamine

MDMA

methylenedioxyphenethylamine (ecstasy)

MDTL

Motherisk Drug Testing Laboratory

MLT

medical laboratory technologist

MS/MS

tandem mass spectrometry

OACAS

Ontario Association of Children’s Aid Societies

OLA

Ontario Laboratory Accreditation

ONCA

Ontario Court of Appeal

OPFPU

Ontario Pediatric Forensic Pathology Unit

PCP

phencyclidine (angel dust)

RRC

Review and Resource Centre

SickKids

Hospital for Sick Children

SOFT

Society of Forensic Toxicologists

SoHT

Society of Hair Testing

SPE

solid-phase extraction

THC

tetrahydrocannabinol

TIAFT

The International Association of Forensic Toxicologists

UKIAFT

United Kingdom and Ireland Association of Forensic Toxicologists

USDTL

US Drug Testing Laboratory

Glossary

alcohol markers; alcohol biomarkers

Substances directly or indirectly linked to previous use of or exposure to alcohol that are measured in hair and other biological samples and used as indicators of chronic use of alcohol.

banding

Term used to describe the expanding region of drug-positive hair as the hair grows away from the surface of the skin as a result of the varying growth rates of individual hairs.

bulk analysis

The preparation of a lock of hair that is not cut into multiple shorter segments before analysis. (In contrast, a longer single length of hair is tested as one sample.) Also refers to the process of assaying a number of different samples in one batch.

calibrators

Standards of increasing drug concentration used to calculate the concentration of drugs present in hair samples.

cross-reactivity (in reference to immunoassay screening tests)

The cross-reactivity of a drug is its ability to cause a response when compared with the target drug (against which the antibody is raised) of the same concentration.

cut-off

A drug concentration above which a sample is classified as positive and below which it is classified as negative. Cut-off concentrations are also used when measuring alcohol markers in hair to assess abstinence or chronic excessive consumption.

dose–response relationship

The measurement of the relationship between the quantity (dose) of a drug administered and the response it produces; or the resulting concentration measured in hair.

ELISA

Enzyme-linked immunosorbent assay: a type of immunoassay screening test that is used to detect the presence of antigens (or drugs) in biological specimens.

ethyl glucuronide (EtG)

A minor metabolite of ethanol and direct alcohol biomarker measured in hair to assess the degree of alcohol consumption over a period of time.

fatty acid ethyl esters (FAEEs)

A group of approximately 20 minor metabolites formed from the use of ethanol, of which some are used as biomarkers to assess chronic excessive alcohol consumption in hair.

gas chromatography–mass spectrometry (GC-MS)

The technique of combining gas chromatography separation with mass spectrometry. It is routinely used to identify drugs and their concentration in hair and other biological samples.

hair colour bias

The natural colour of hair can affect the absorption of some drugs and, in turn, the interpretation of the degree of exposure to drugs.

homogenization

In the context of hair analyses: where very small segments of hair, or powdered hair, are mixed to ensure a uniform sample of hair that can be used to prepare for an analysis.

immunoassay

A biochemical test that detects the presence of the target compound through the use of antibodies.

internal standard, deuterated

A substance (usually a drug or metabolite) in which some hydrogen atoms are replaced by the hydrogen isotope deuterium, which is added in known amounts to specimens (hair extracts) to correct for variability in the isolation and detection techniques.

inter-subject variation

In the context of hair analyses: variability among individuals in the drug content in hair or in the effect of drugs – even when those individuals are exposed to identical amounts of a drug.

Levey-Jennings chart

A plot of drug concentration obtained from a laboratory control sample for each consecutive analysis with limits reflecting 1, 2, and 3 standard deviations from the average. Used to assess performance of an analytical assay.

liquid chromatography–tandem mass spectrometry (LC-MS/MS)

The technique of combining a high-performance liquid chromatography separation with tandem mass spectrometry. It is routinely used to identify drugs and their concentration in hair and other biological samples.

mass-to-charge ratio (m/z)

The ratio of the ionic weight of an ion produced in a mass spectrometer over the number of charges on the ion.

meconium

The first fecal matter produced by a newborn.

metabolite

A substance produced from the parent drug during its transit through the body. It is produced through one or more types of metabolic processes.

qualitative (result)

A result that provides an indication of drug presence as a positive (yes) or negative (no), rather than providing actual quantitative results.

quantitative (result)

A result that is reported containing the actual concentration of the drug in a sample, indicating a degree of precision and accuracy.

reagents

Materials, usually solutions of buffers and/or other chemicals, used in the conduct of analytical assays.

reference laboratory

A laboratory that is used to verify another laboratory’s result; usually one with accredited and properly validated methods and known to produce reliable and reproducible results.

segmental analysis

The process of cutting a strand or a length of hair into smaller sections with the aim of analyzing the content of each section (segment) and obtaining some history of relative exposure to drugs.

Chapter 1
Executive Summary

Introduction

From the late 1990s to April 2015,1 the Motherisk Drug Testing Laboratory at Toronto’s Hospital for Sick Children held itself out as a leader in the field of hair testing for drugs of abuse. The Motherisk Drug Testing Laboratory’s (MDTL’s or the Laboratory’s) main customers were Ontario child protection agencies seeking to determine if a parent or caregiver had used drugs or alcohol. Between 2005 and April 2015, the Laboratory tested hair samples from more than 16,000 individuals at the request of Ontario child protection agencies. The child protection agencies expected that MDTL, which was housed at the prestigious Hospital for Sick Children, was providing adequate and reliable expert opinions to them to use in making decisions involving issues of child protection. In fact, however, the hair test results and the opinions delivered by MDTL were neither adequate nor reliable.

Society places great value on the parent–child relationship. The law protects that relationship and the right of parents to nurture and care for their child. At the same time, it places the greatest importance on the safety and well-being of the child.

In Ontario, child protection services are governed by the Child and Family Services Act (CFSA). The CFSA’s paramount purpose is to promote the best interests, protection, and well-being of children. In addition, the CFSA recognizes that, although parents may need help in caring for their children, that help should, wherever possible, support the autonomy and integrity of the family unit, be provided on the basis of mutual consent, and be the least disruptive course of action available and appropriate to help a child. Although child protection agencies offer support for families, provide care for children, and help families to stay together, they may, in extreme cases, remove children from their family’s care.

Parents interacting with child protection agencies are often among the most vulnerable or marginalized members of our society. They are frequently dealing with multiple challenges, including poverty, inadequate housing, mental and other health concerns, and, sometimes, substance abuse.

When a child is removed from parental care, that interference with the parent–child relationship causes great distress, both for children and for their parents (or other caregivers). Lost is the day-to-day interaction that is central to family life. For that reason, it is imperative that the evidence relied on by child protection agencies when seeking to remove children from parental care is adequate and reliable.

The Laboratory and the Independent Review

MDTL was one facet of the Hospital’s Motherisk Program, which provides information and guidance to members of the public and to physicians about the potential risks to a developing fetus or infant from exposure to drugs, chemicals, diseases, radiation, and environmental agents. MDTL began as a research laboratory, carrying out cutting-edge research on neonatal hair analysis. By the late 1990s, MDTL was receiving an increasing number of requests from child protection agencies to test hair samples for drugs of abuse. In 2001, the Laboratory began to promote its hair-testing services to child protection agencies through a variety of presentations and seminars.

Several individuals held key responsibilities at the Laboratory:

  • Dr. Gideon Koren, a physician and clinical toxicologist, was the founder and laboratory director of MDTL, and the person with ultimate responsibility for the Laboratory;
  • Julia Klein was the laboratory manager until April 2005, when her employment was terminated; and
  • Joey Gareri became the laboratory manager after Ms. Klein’s departure.

In 2005, MDTL tested more than 1,500 samples at the request of child protection agencies, and that number continued to increase over the ensuing years. From time to time, MDTL tests were also used in criminal cases, one of which led directly to this Independent Review. On October 14, 2014, the Court of Appeal for Ontario allowed the appeal of Tamara Broomfield from two criminal convictions on charges that she had administered cocaine to her two-and-a-half-year-old child over a 14-month period. Ms. Broomfield had been convicted, in part, on the basis of evidence from Dr. Koren about the results of MDTL hair tests performed on her child.

The Court of Appeal admitted fresh evidence from Dr. Craig Chatterton, the deputy chief toxicologist in the Office of the Chief Medical Examiner of Alberta, who criticized MDTL’s hair-testing methodology and its interpretation of the hair test results. The Court of Appeal concluded that there was a “genuine controversy” about the science and the methodology used by the Laboratory and quashed two of Ms. Broomfield’s cocaine-­related convictions. The Court of Appeal ordered a new trial. In keeping with the joint submission of the Crown and the defence, however, the Court stayed the new trial as not being in the interests of justice for several reasons, including that Ms. Broomfield had already served the equivalent of a 49-month sentence.

Just over a month later, on November 26, 2014, the cabinet of the Government of Ontario established this Independent Review and appointed me the Independent Reviewer. On April 22, 2015, the mandate of the Independent Review was expanded and directed me to conduct a review and provide a report of my findings and recommendations respecting

  • the adequacy and reliability of the hair-strand drug and alcohol testing methodology utilized by MDTL between 2005 and 2015 for use as evidence in child protection and criminal proceedings; 
  • the extent to which the operation of the MDTL laboratory between 2005 and 2015 was consistent with internationally recognized forensic standards;
  • other matters related to the operation of the MDTL laboratory that I consider necessary and appropriate to address as a result of my review; and
  • whether the use of evidence derived from MDTL’s hair-strand drug and alcohol testing in criminal and child protection proceedings has implications warranting an additional review or process with respect to specific cases or classes of cases and, if so, the nature and extent of any such review or process.
  • For the reasons contained in this report, I make the following findings:
  1. The hair-strand drug and alcohol testing used by MDTL between 2005 and 2015 was inadequate and unreliable for use in child protection and criminal proceedings.
  2. Between 2005 and 2015, MDTL operated in a manner that did not meet internationally recognized forensic standards.
  3. The Hospital for Sick Children did not provide meaningful oversight over MDTL.
  4. The use of MDTL hair-testing evidence in child protection and criminal proceedings has serious implications for the fairness of those proceedings and warrants an additional review.

Use for Forensic Purposes

MDTL’s hair tests were forensic in nature, and the service it offered to child protection agencies and law enforcement was a forensic one. A hair test can be forensic even if it is never tendered as evidence and even if no court proceeding is ever initiated. What distinguishes a clinical test from a forensic test is the purpose behind the test. If the test is either carried out or used for a legal purpose, it is a forensic test.

To be used in child protection or criminal proceedings, forensic tests must be carried out in accordance with forensic standards. International forensic standards are articulated in several sources:

  • international guidelines on the science of hair testing (including the consensus statements of the Society of Hair Testing), as well as on the standards of forensic toxicology laboratories generally (including, for example, the guidelines of the Society of Forensic Toxicologists and the American Academy of Forensic Sciences);
  • the standards that are required to be met for the accreditation of forensic laboratories (including the requirements under ISO 17025:2005, General requirements for the competence of testing and calibration laboratories);
  • best practices that are commonly accepted and recognized by forensic toxicology laboratories around the world; and
  • the academic literature on the science of hair testing and the interpretation of hair test results.

Although the leaders of MDTL had relevant experience as research or clinical toxicologists, none of them had any formal training or experience in forensic toxicology. Indeed, none of them considered themselves to be forensic toxicologists. Perhaps this lack of training and experience is why neither MDTL nor the Hospital for Sick Children (SickKids or the Hospital) appears to have appreciated that the Laboratory was engaged in forensic work and that it was required to meet forensic standards. The result was inevitable: MDTL’s testing and operations fell woefully short of internationally recognized forensic standards.

Adequacy and Reliability

The hair-strand drug and alcohol testing used by MDTL between 2005 and 2015 was inadequate and unreliable for use in child protection and criminal proceedings. I explain these conclusions below with reference both to MDTL’s analytical practices and to its interpretation and communications practices.

Analytical Practices: 2005–10

In the 2005–August 2010 period, MDTL’s analytical practices for drugs of abuse suffered from two fatal flaws:

  1. the Laboratory relied on the unconfirmed results of its enzyme-linked immunosorbent assay (ELISA) tests2 – a preliminary screening test – both qualitatively (to distinguish positive hair samples from negative ones) and quantitatively (to calculate the drug concentration in the sample); and
  2. the Laboratory had no written standard operating procedures for the hair tests it carried out, thereby calling into question the reliability and standardization of all its testing procedures.

Although there were other deficiencies, these two alone were sufficient to render MDTL’s hair tests inadequate and unreliable during this period.

ELISA Tests

From January 2005 to August 2010, MDTL tested all hair samples for drugs of abuse using ELISA. The particular ELISA test that MDTL used came in the form of a kit supplied by Immunalysis Corporation (Immunalysis), a well-known and well-respected company in California. The drugs and metabolites for which MDTL tested most frequently during this period were cocaine, benzoylecgonine, and cannabis / THC.

ELISA is widely used in both forensic and clinical toxicology as a screening test. It is intended to determine quickly if a sample is negative and merits no further testing or if it is a preliminary positive, in which case the sample must be tested again using another method (a confirmation test) to determine if the sample is in fact positive. The requirement to carry out a confirmation test on any preliminary positive results from immunoassay-based screens, such as ELISA, was highlighted in the literature and in all internationally recognized hair-testing standards well before 2005. Indeed, the Immunalysis kits that MDTL used included an explicit warning for the user about the preliminary nature of the ELISA results.

Until August 2010, MDTL reported its ELISA results to customers without a confirmation test. It reported its ELISA results – both qualitatively (as positive vs. negative) and quantitatively (in the form of a numerical drug concentration for positive samples). Such a practice is inconsistent with internationally recognized forensic standards and is unacceptable for two important reasons.

First, unlike confirmation techniques such as gas chromatography–mass spectrometry (GC-MS) or liquid chromatography–tandem mass spectrometry (LC-MS/MS), ELISA cannot identify the substances within a sample. Second, MDTL’s practice of using ELISA to quantify drug concentrations is simply unheard of in forensic toxicology laboratories. No forensic toxicology laboratory in the world uses ELISA testing the way MDTL did.

Absence of Standard Operating Procedures and Contemporaneous Documentation

Both forensic and clinical laboratories routinely have standard operating procedures in place for each of the analytical tests they perform. In addition, forensic laboratories are required to maintain contemporaneous documents identifying what steps were taken in respect of a sample, who took them, when and why, as well as any decisions made that might be relevant to the result. This contemporaneous documentation is essential to permit meaningful review of the reliability of the test results.

MDTL was not able to produce any written analytical procedures that were in use for its drug tests from 2005 to 2010. In addition, from 2005 to 2010, MDTL did not maintain contemporaneous documentation of the steps its laboratory personnel performed for a specific sample, including records of who performed which step in the process, what was performed, or when or how it was performed. MDTL’s practices were unacceptable and fell well below expected standards for a forensic laboratory.

Additional Flaws

In addition to these critical flaws in MDTL’s hair-testing process, other aspects of the Laboratory’s analytical methods fell short of internationally recognized forensic standards. For example:

  • There was no oversight of the laboratory technician who was assigned to carry out the ELISA tests, with the result that there were significant reporting anomalies or errors in the case files reviewed by the Independent Review.
  • Staff failed to wash hair samples routinely before analysis.
  • The chain-of-custody procedures were inadequate.
  • The record keeping was inadequate.

Analytical Practices: 2010–15

In September 2010, MDTL began routinely to confirm the majority of its test results for drugs of abuse, initially using GC-MS and, after May 2014, LC-MS/MS. MDTL also improved many of its laboratory practices by adopting written standard operating procedures and a quality management system. However, even in the September 2010–April 2015 period, MDTL continued to fall short of internationally recognized forensic standards by

  • relying on a flawed GC-MS procedure;
  • continuing to rely on unconfirmed ELISA results for certain drugs and for neonatal samples (though MDTL stopped reporting the quantitative results generated by those tests);
  • continuing to test unwashed hair samples from children and neonates; and
  • not maintaining appropriate chains of custody of the samples it tested.

Analytical Practices: Alcohol Markers

Although MDTL’s hair-testing methodology for the alcohol markers known as fatty acid ethyl esters (FAEEs), at least from 2007 to April 2015, did not suffer from the same deficiencies as its testing methods for drugs of abuse, MDTL’s FAEEs tests were still inadequate and unreliable for forensic purposes: first, because of deficiencies in its GC-MS procedures; and second, because of its misapplication of cut-offs to assess alcohol consumption. In addition, MDTL used non-standard hair lengths in its tests. The Laboratory also relied on FAEEs analysis alone to assess alcohol consumption – a practice that was acceptable at the time, but is now considered by many hair-testing experts to be unreliable, given the evolution in the science.3

2005–15: Interpretation and Communication Practices

The effective interpretation and communication of results are essential elements of forensic toxicology. It is critical that hair test results not be misinterpreted or over-interpreted and that the limitations on any interpretation be clearly conveyed to the recipient of the result. Forensic toxicologists are charged with ensuring that toxicology results are communicated effectively to users of the results, including the justice system.

MDTL’s interpretation practices and its communications to users were inadequate and further undermined the reliability of the results that the Laboratory provided for use in child protection and criminal cases.

Misuse of Concentration Ranges

MDTL generated “concentration ranges” that purported to be able to identify concentrations for drugs and metabolites in a hair sample as “very low” (or “trace”), “low,” “medium,” “high,” and “very high.” MDTL relied heavily on these concentration ranges to interpret its test results for drugs of abuse and to communicate the meaning of those test results to its customers. Using concentration ranges in this way was unacceptable for two reasons:

  • Concentration ranges do not account for inter-subject variation: two people who use the same amount of a drug are not expected to have the same concentration of drug in their hair because of differences such as in metabolism and hair characteristics.
  • MDTL’s concentration ranges were built on flawed data consisting of unconfirmed ELISA results, which are inherently unreliable, and data that pooled together results from both washed and unwashed samples.

Although MDTL appeared to be aware of the dangers of using concentration ranges to interpret hair test results, it nevertheless used them to draw inappropriate and incorrect conclusions about an individual’s level and frequency of drug use, using such terms as “repeated,” “isolated,” or “frequent” to describe how often a person used a drug; and “bingeing” or “intensive use” to describe the quantity.

Poor Communication to Customers

MDTL’s poor communications with customers went beyond its misinterpretation and over-interpretation of its tests results. MDTL inadequately communicated its test results to customers for several reasons:

  • No one at MDTL had the expertise to provide a forensic toxicology interpretation of the hair test results.
  • MDTL regularly sent the bare test results to its customers without providing an interpretation of the data – an essential component of a meaningful report.
  • When MDTL did interpret the test results, staff often provided the interpretation over the telephone and not in writing.
  • MDTL made repeated interpretation errors when reporting on its results for cocaine (and its metabolites) and cannabis / THC, two of the most common drugs for which the Laboratory tested.

Proficiency Testing and Accreditation

When questions were raised about MDTL’s analytical methods both on the Broomfield appeal and in its aftermath, MDTL and the Hospital relied on the Laboratory’s performance in proficiency tests as evidence of the robustness of its analytical methods. They should not have done so. For the most part, these proficiency test results were not based on test results that were actually generated by the Laboratory, and MDTL’s participation in proficiency testing therefore provided no evidence of the reliability or robustness of the methods used by the Laboratory.

Likewise, MDTL relied on its Ontario accreditation as a clinical laboratory as evidence of the quality and reliability of its hair-testing methodology. However, this accreditation process did not scrutinize or validate the adequacy and reliability of MDTL’s hair tests or the robustness of its analytical methods and interpretations.

Proficiency Testing

Proficiency testing is an external assessment of the quality of a laboratory’s results by means of an inter-laboratory comparison. From 2001 to 2014, MDTL participated in proficiency testing offered by the Society of Hair Testing (SoHT), which is the only body that offers proficiency testing for hair-testing laboratories. From 2011 to 2014, MDTL also participated in the proficiency testing scheme that the SoHT offered for fatty acid ethyl esters (FAEEs).

Despite MDTL’s emphasis on its proficiency test results from 2001 to 2009 to support its use of ELISA, the quantitative results that MDTL submitted annually to the SoHT during this period were not actually generated by the Laboratory. Rather, MDTL frequently submitted quantitative results that had been produced by another laboratory or results that were inconsistent with its own data. As a result, the proficiency tests that MDTL claimed demonstrated the reliability of its ELISA tests did not assess the proficiency of MDTL’s ELISA tests at all. Accordingly, they provide no assurance about the reliability of MDTL’s ELISA tests during those years.

After MDTL changed its testing methodology in 2010, with one exception it began to submit to the SoHT its own GC-MS–generated data (and, in 2014, its LC-MS/MS–generated data). Although MDTL’s proficiency test results from 2010 to 2014 did not yield any false positive or false negative results (with one exception), in some years its quantitative results deviated significantly from those of the SoHT’s reference laboratories and other participating laboratories.4

Accreditation

In Ontario, accreditation (including participation in proficiency testing) is mandatory for clinical laboratories, but is not a requirement for laboratories providing forensic services. If a laboratory can otherwise demonstrate that it adhered to internationally recognized forensic standards in its processes and procedures, the fact that it lacks accreditation is not determinative of whether its test results may be admissible in court or are otherwise reliable for forensic purposes.

MDTL’s participation in SickKids’s clinical laboratory accreditation process does not demonstrate the adequacy and reliability of MDTL’s tests for forensic purposes for three reasons. First, MDTL was not, and has never been, accredited to provide forensic services; rather, the accreditation program in which it participated was clinical in nature. Second, MDTL did not participate in its first accreditation inspection until January 2011; accordingly, the fact of MDTL’s accreditation does not support the reliability of its hair tests before that time. Third, the accreditation process in which MDTL participated did not assess the robustness of either MDTL’s analytical methods or the interpretations it provided to users of its test results.

Ultimately, although MDTL made significant improvements to both its analytical methods and its quality management system in the lead-up to its accreditation inspection in January 2011, and although the Ontario Laboratory Accreditation program helped to usher in a new era for MDTL, the fact that the Laboratory passed its first accreditation inspection under this program in January 2011 (and a second time in December 2014) is not evidence that MDTL’s testing methods or interpretations were adequate or reliable for forensic purposes in the years following accreditation. As explained above, they were neither.

At No Time Did MDTL Meet Forensic Standards

In conclusion, for the reasons set out above, the Laboratory never met internationally recognized forensic standards. Indeed, MDTL’s leadership failed to appreciate that the Laboratory was doing forensic work. None of the Laboratory’s leadership had any formal training in forensic toxicology.

The Laboratory, nevertheless, actively marketed its services to child protection agencies. Throughout the 10-year period under review, MDTL was testing an average of approximately 2,000 hair samples each year for child protection agencies. From time to time its hair tests were also used in criminal proceedings. Indeed, Dr. Koren, Ms. Klein, and Mr. Gareri all testified in court as expert witnesses. Dr. Koren and Mr. Gareri both gave evidence in the Broomfield case. Given the manner in which MDTL operated, Dr. Koren, Ms. Klein, and Mr. Gareri should have appreciated that MDTL was carrying out forensic, not clinical, work.

Oversight Failures at the Hospital for Sick Children

MDTL was housed in one of the most prominent and well-respected children’s hospitals in Canada – the Hospital for Sick Children. SickKids has a sophisticated infrastructure of oversight and accountability for its laboratories, yet no one charged with overseeing MDTL provided any meaningful oversight. What went wrong at the Hospital?

MDTL was originally a research laboratory, housed in the Research Institute at SickKids. The Hospital told the Independent Review that it was not until 2005 that it became aware of the nature and quantity of the testing that the Laboratory was carrying out for non-research purposes. However, even when the Hospital’s leadership became aware that MDTL was routinely carrying out hair tests for child protection cases, the Hospital viewed MDTL’s work as being a shift from research toward clinical practice. There is no evidence that anyone in a leadership position at the Hospital meaningfully considered the reality that MDTL was, for the most part, carrying out hair tests for forensic, not clinical, purposes and that it needed to meet forensic standards.

When, in 2005, it learned of the extent of the tests that MDTL was performing for non-research purposes for child protection agencies, the Hospital considered bringing MDTL into SickKids’s clinical licensing and accreditation process. Even so, MDTL was included in neither the 2006 Ontario Laboratory Accreditation inspection of the Hospital’s clinical laboratories nor the 2008 Ontario Laboratory Accreditation midterm self-assessment of the Hospital’s clinical laboratories.

In fact, in 2008, when the Department of Paediatric Laboratory Medicine (DPLM) was performing its midterm self-assessment, it specifically turned its mind to MDTL’s inclusion in that process. However, the Hospital decided not to include MDTL because it recognized that accreditation would likely not be achieved. This decision is not surprising. If MDTL had been included in the midterm self-assessment at the time, it would have been clear to the accrediting body that MDTL did not meet the requirements of the Ontario Laboratory Accreditation program for a clinical laboratory. The stark reality is that, at this time, MDTL did not have a quality management system or any standard operating procedures for its hair tests for drugs of abuse, both of which were basic requirements under the Ontario Laboratory Accreditation program.

Even when the Hospital eventually brought MDTL into the Ontario Laboratory Accreditation process, it took no steps to ensure that MDTL’s hair tests were fit for their intended purpose or that MDTL had the appropriate equipment, infrastructure, personnel, and expertise to carry out hair tests for drugs of abuse and alcohol markers. None of the programs, departments, or divisions within the Hospital ever took ownership of MDTL as a laboratory, and no clear lines of accountability were ever drawn. The result was inevitable: MDTL slipped through the cracks. The Hospital exercised no meaningful oversight over the Laboratory.

My conclusion that SickKids failed to exercise meaningful oversight over MDTL’s work must be considered in the context of the Hospital’s experience with Dr. Charles Smith, a pediatric pathologist who worked at SickKids. Dr. Smith was the first director of the Ontario Pediatric Forensic Pathology Unit (OPFPU), which was housed in the Hospital. In April 2007, the Government of Ontario appointed the Honourable Justice Stephen T. Goudge as commissioner of the Inquiry into Pediatric Forensic Pathology in Ontario. The Hospital was a party with standing at the inquiry. Commissioner Goudge released his report on October 1, 2008 (Goudge Report) – the same year in which the Hospital determined that clinical accreditation might not be achieved for MDTL. The Goudge Report is relevant to this Independent Review for at least three reasons:

  1. It highlighted the dangers associated with having a laboratory within the institution that routinely provided a forensic service yet was led by individuals who lacked any forensic training.
  2. It concluded that the Hospital’s lines of oversight and accountability over the forensic pathology service lacked clarity and created a vacuum where nobody was held accountable for the forensic pathology service.
  3. Commissioner Goudge noted the role that SickKids’s reputation for excellence played in positioning Dr. Smith as a leading expert in his field, notwithstanding his lack of forensic expertise.

All three of these lessons should have been applied to MDTL, but the Hospital did not do so.

Child Protection and Criminal Proceedings

MDTL test results were employed at each stage of child protection proceedings, depending on the practice of the particular agency. As submissions from the Family Lawyers Association and the Aboriginal Legal Services of Toronto described the situation, MDTL tests were used

  • to confirm suspicions of drug and alcohol use;
  • to obtain an accurate level of use;
  • to monitor levels of drug and alcohol use over time and assess parents’ compliance with terms and conditions for access to a child;
  • as significant evidence of a caregiver’s drug or alcohol use, or the exposure of children to drug use;
  • to encourage a parent or caregiver to consent to agency intervention, including a temporary care order;
  • to test a caregiver’s credibility; and
  • as a term of a court order.

In some cases, MDTL hair test results were used as a measure of a parent’s credibility in circumstances in which the test results did not accord with the parent’s self-reporting of drug or alcohol use.

The Family Lawyers Association expressed particular concern about the use of MDTL hair test results at the early stages of intervention, saying that a positive hair test at that point might “be a significant and almost incontrovertible piece of evidence” in part because “it affects the tenor and trajectory of the case.” As a result, a child might remain in foster care for some time, and at the next hearing the parent would also be faced with “a disadvantageous status quo.”

It is noteworthy that child protection agencies regularly introduced MDTL’s hair test evidence in court through affidavit, often by a child protection worker, attaching MDTL results and interpretation reports as exhibits. Some affidavits simply attached a results report (with no interpretation) or noted the fact of a positive test, along with the quantitative result with reference to the concentration ranges, in the body of their affidavit. Other affidavits recounted the content of a telephone consultation with an MDTL laboratory counsellor, who was not qualified to give a forensic opinion. In short, MDTL interpretations were commonly admitted as evidence on the basis of the child protection worker’s information and belief at all stages of child protection proceedings: temporary care hearings, status review hearings, and even sometimes on summary judgment motions.

Although these results and interpretations were expert opinion evidence, they were infrequently recognized or treated as such by child protection counsel, parents’ and children’s counsel, or the court. They were often introduced into evidence on consent. As a result, the MDTL evidence was rarely tested against the admissibility requirements for expert opinion evidence that apply in civil proceedings, including child protection proceedings. Accordingly, courts frequently accepted the test results and the concentration ranges as a reliable measure of use – with little, if any, examination of the forensic qualifications of the MDTL representative who communicated those results to the child protection agency, and without any analysis of the adequacy and reliability of the test methodology employed by the Laboratory.

However, in the vast majority of cases MDTL hair test results were only one piece of the evidence that was available to assess a child’s safety and risk. As one child protection agency put it, “several other factors, such as the state of the home, the demeanor of the parents, the parents’ overall health, the child’s functioning, attachment and development were weighed in decision making.”

A review of some reported child protection cases involving an MDTL test result reveals examples where it appears that MDTL test results were but a small factor in the outcome, and other examples where it appears that the results were given significant weight. Given my conclusion that the hair-strand drug and alcohol testing used by MDTL between 2005 and 2015 was inadequate and unreliable for use in child protection and criminal proceedings, the use of that evidence has serious implications for the fairness of those proceedings and warrants a further review of individual cases or classes of cases.

In addition to their extensive use in child protection proceedings, hair test results from the MDTL were also used as evidence in criminal proceedings. However, in contrast to the hair samples from the 16,000 and more individuals tested by MDTL between 2005 and 2015 at the request of child protection agencies, few hair samples were processed by the Laboratory for criminal matters. The information provided to the Independent Review by the Crown Law Office – Criminal identified six cases, including Broomfield, in which MDTL evidence was used by the Crown and resulted in a conviction. Criminal cases involving MDTL hair-testing evidence also merit the scrutiny and resources of a further review.

Summary of Recommendations

The Province of Ontario should, as expeditiously as possible, establish a Second Review of those individual cases that may have been affected by MDTL’s flawed hair-testing methodology.

The Second Review should be carried out by a Commissioner, a sitting or retired judge, appointed under the Public Inquiries Act with the power to

  • summons witnesses, as well as documents and other items relevant to the subject matter of the Second Review; and
  • access all files appropriate to the criminal or child protection context, including court files, child protection files, exhibits, and transcripts, in order to fulfill the role of assessing and triaging cases.

The Commissioner should lead a Review and Resource Centre (RRC), which would provide support to potentially affected persons (including parents, children, young adults, siblings, and adoptive parents) involved in past child protection proceedings who may have been affected by MDTL test results. This support would provide efficient, comprehensive access to the resources necessary to permit them to make an informed decision about any steps that might be available and appropriate. These resources should include, at no charge,

  • appropriate counselling assistance;
  • appropriate legal advice from lawyers listed on the RRC roster;
  • alternative dispute resolution services, including mediation; and
  • fresh parenting assessments to assess if a parent currently has the capacity to have a relationship with a child.

The Province of Ontario should ensure that the RRC has sufficient resources to provide meaningful assistance and to fulfill its mandate. The Province should also ensure that other child protection and justice-system institutions that must deal with the implications of this Report and the Second Review have sufficient resources to do so.

The Commissioner should provide widespread public notice of the Second Review in the manner that is most likely to come to the attention of those persons potentially affected by MDTL tests. Any parent who had an MDTL test and has a child who is no longer in her or his care, and where that child has not yet been placed for adoption or the adoption has broken down, should receive personal notice of the Second Review. In addition, any potentially affected person should be able to initiate a request for a review.

With the support of the Ministry of Children and Youth Services, child protection agencies should immediately identify any cases involving MDTL hair test results that remain open. In those cases, and in cases where a child has not yet been placed for adoption, child protection agencies should

  • contact the parents or their lawyers to advise them of the potentially flawed hair test results and the creation of the Second Review;
  • assess these cases without regard to MDTL test results unless and until those results are confirmed, if they can be; and
  • provide a complete copy of the unredacted file to the RRC as soon as possible.

Child protection agencies should ensure that no person or organization relies on any MDTL test results for any purpose in any current or future proceeding.

The Commissioner should consider all cases where there was an MDTL test result and where one of the terms of an operative court order requires the parent to comply with ongoing hair testing.

The Commissioner should work with children and youth, including youth who have experienced the child protection system, to ensure that their voices, both individually and collectively, are heard throughout the Second Review.

The RRC should have the capacity to assist residents across Ontario. It should make effective use of its online presence and telephone communication, including an immediately available toll-free number and e-mail address answered by well-trained and knowledgeable staff.

If an affected child has been placed for adoption (particularly without an openness order) in part because of flawed evidence from MDTL, the currently available range of remedies is legislatively limited. I recommend that the Commissioner consult broadly with affected persons, including youth, adoptive parents, and other caregivers, as well as with experts in the area, in an effort to develop approaches to resolution based on the best interests of the child. With this opportunity for broad consultation, and once the circumstances of individual cases or classes of cases have been identified, the Commissioner will be in a position to identify the parameters of the problem and any public policy considerations at issue.

I recommend that the Commissioner also consider whether, in appropriate circumstances, birth parents or other affected individuals should be provided with the ability to file information on the adoption registry or in the child protection file concerning the MDTL controversy and the possible role of flawed hair tests in affecting the outcome of a child protection proceeding. This information would then be available for affected children who access the registry.

Public confidence in the justice system demands that any evidence relied on in criminal cases be adequate and reliable for forensic purposes. MDTL’s hair tests did not meet this standard. Individuals who may have been affected by an MDTL test result in a criminal proceeding should also have access to the services of the RRC. With regard to individuals who may have been affected by an MDTL test result in a criminal proceeding:

  • They should be permitted to access the counselling services, appropriate legal advice, and guidance from the RRC on the next steps they may wish to take.
  • In cases in which an individual seeks to set aside convictions based on alleged errors in the Laboratory’s work, the Crown Law Office – Criminal should assist in expediting the convicted person’s access to the Court of Appeal and in facilitating a determination of the substantive issues in the cases, unencumbered by unnecessary procedural impediments. Such assistance should include
    • consenting to defence applications for an extension of time within which to appeal;
    • working toward agreement with the defence on evidentiary or procedural protocols for applications to extend the time within which to appeal or for introducing fresh evidence on appeal or respecting the appeal itself; or
    • narrowing the issues that need to be resolved by the Court.

The Province of Ontario, either directly or through Legal Aid Ontario, should adequately fund potentially affected persons to access legal and forensic opinions and advice for the purpose of pursuing either an appeal or a new trial or any other legal remedy, including a section 696.1 Criminal Code application, to set aside an existing criminal conviction. In appropriate cases, this funding should extend to obtaining an expert opinion to determine whether or not there is merit to such an appeal or application.

The order in council establishing my Review did not provide me with a mandate to make recommendations about individual compensation. Accordingly, I do not make any recommendations concerning compensation. Questions of any monetary compensation are left to the Second Review, the Province of Ontario, or civil litigation for consideration and determination.

The Hospital for Sick Children should carefully examine its programs to determine which, if any, are offering forensic services and should exercise meaningful oversight over any such services. The Hospital must ensure that appropriate standards and training are in place for any of its staff who testify in court as experts. Ultimately, the Hospital will be held accountable for the quality of any forensic services it provides.

There should be ongoing and expanded continuing legal education for counsel, including child protection counsel, and for the judiciary regarding expert evidence, particularly expert scientific evidence. Members of the bar should ensure that they understand the scope and limitations of an expert’s opinion and expertise. In addition, they should exercise care to qualify an expert properly and set precise parameters on an expert’s area of expertise. Counsel should be vigilant to ensure that justice is not undermined by the use of flawed forensic evidence. The recommendations of the Goudge Report continue to provide valuable insights on these issues.

Conclusion

The citizens of Ontario are justifiably proud of the Hospital for Sick Children, which is one of the world’s leading children’s hospitals. I am confident that the Hospital will reflect deeply on what went wrong at MDTL and within its own institution, particularly because it appears that the Hospital has not adapted and implemented the lessons to be learned from the Goudge Report. There should never need to be another review or inquiry into how the Hospital, its professional staff, or its programs interact with the justice system.


Notes

1 On April 17, 2015, the Hospital shut down the Motherisk Drug Testing Laboratory for all non-research purposes.

2 A glossary of scientific terms appears at the beginning of the Report.

3 MDTL did not test hair samples for ethyl glucuronide (EtG). From February 2011 until the Laboratory closed, it sent all samples that fell above the Society of Hair Testing cut-off for FAEEs for EtG tests to the US Drug Testing Laboratory (USDTL) in Des Plaines, Illinois. The Independent Review has not assessed the USDTL hair-testing methodology, but MDTL also misapplied the SoHT cut-offs with respect to the EtG results it received from USDTL.

4 The exception was in 2010, when MDTL reported the results of its reference laboratory for cannabis / THC, which yielded a false negative.

Chapter 2
The Independent Review:
Origins, Mandate, and Process

1. Introduction

1. This Independent Review was initiated by the Province of Ontario, in response to controversy concerning the scientific reliability for forensic purposes of hair tests performed by the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) at the Hospital for Sick Children (SickKids or the Hospital). That controversy arose in the context of a criminal appeal to the Court of Appeal in R v Broomfield.5

2. This chapter will briefly address the issues identified in Broomfield, describe the original and expanded mandate for the Independent Review, explain how the Independent Review was set up, and then describe its process. In describing the process, I acknowledge those who contributed so greatly to ensure both a comprehensive and efficient review and the timely release of this Report, which is important for the potentially affected individuals. Finally, I touch on the nature of my findings.

2. Origins of the Review: The Broomfield Case and Aftermath

3. On August 1, 2005, Tamara Broomfield rushed her two-and-a-half-year-old child to the local hospital because the child was suffering from seizures. Within hours, the child was transferred to SickKids.6 In addition to identifying a potentially lethal dose of cocaine in the child’s bodily fluids, the Hospital’s medical investigation found that the child had several broken ribs in various stages of healing, as well as a broken wrist.

4. Ms. Broomfield took the position that the cocaine identified in her child’s bodily fluids may have resulted from one accidental exposure. On August 9, 2005, eight days after the child’s initial admission and at the request of the doctors treating the child, the child’s hair was collected and tested by MDTL for drugs of abuse. In December 2005, the child’s hair sample was tested again, this time at the request of a child protection agency.

5. On August 14, 2005, Ms. Broomfield was charged with several criminal counts arising from the injuries suffered by her child. At trial in 2009, the Crown called expert evidence from MDTL’s laboratory director, Dr. Gideon Koren, about the results of MDTL’s hair tests. Dr. Koren testified that the child must have ingested substantial amounts of cocaine over the preceding 14 months. In his opinion, the test results were inconsistent with a single accidental exposure to cocaine or environmental exposure, such as to smoke.7

6. On April 1, 2009, the trial judge convicted Ms. Broomfield of several counts related to the child’s broken bones and her failure to have the fractures treated. The trial judge also accepted Dr. Koren’s evidence and convicted her of administering a noxious substance to her child over a 14-month period (as well as convicting her on a charge arising from the potentially lethal dose of cocaine that led to the child’s hospitalization on August 1, 2005).

7. Ms. Broomfield appealed all her convictions but pursued her appeal only of the convictions based on her child’s ingestion of cocaine. On the appeal, she tendered fresh evidence from Dr. Craig Chatterton, the deputy chief toxicologist in the Office of the Chief Medical Examiner of Alberta. Dr. Chatterton criticized MDTL’s hair-testing methodology, including its use of an unconfirmed immunoassay-based analysis as well as its interpretation of the results.

8. The Court of Appeal admitted the fresh evidence from Dr. Chatterton and concluded not that the hair tests carried out by MDTL were unreliable, but rather that there was a “genuine controversy” about the science and the methodology used. The Court of Appeal quashed the two cocaine-related convictions and ordered a new trial, but, in keeping with the joint submission of the Crown and the defence, stayed the new trial as not being in the interests of justice for several reasons, including that Ms. Broomfield had already served the equivalent of a 49-month sentence.

9. The Court of Appeal decision in Broomfield, which was released on October 14, 2014, sounded an alarm about the hair-testing methodology that MDTL used. Since the late 1990s, MDTL had held itself out as a leader in the field of hair testing for drugs of abuse. MDTL’s main customers were child protection agencies seeking hair test results for use in child protection matters (usually to determine if a parent or caregiver had used or been exposed to drugs of abuse over a period of time). However, as is evident from Broomfield, the Laboratory also occasionally performed tests for criminal cases. In the aftermath of the Broomfield appeal, the public, legislators, family lawyers and criminal lawyers, child protection organizations, and others raised questions about the controversy surrounding MDTL’s hair testing. For its part, the Hospital acknowledged the need for the province to undertake a review of MDTL’s past use of its immunoassay-based testing methodology.

10. This issue of potentially flawed scientific evidence was reminiscent of an earlier public inquiry. The forensic use of scientific evidence had received attention in relation to the forensic pathology evidence of a pediatric pathologist at the Hospital for Sick Children, Dr. Charles Smith, and the role of that evidence in certain criminal convictions. In 2007, Ontario appointed the Honourable Justice Stephen T. Goudge to act as commissioner for the Inquiry into Pediatric Forensic Pathology in Ontario (Goudge Inquiry). His 2008 report detailed how flawed pediatric forensic pathology evidence can affect the criminal justice system.8 In addition, Commissioner Goudge grappled with the difficulties that the justice system in Ontario (and in other jurisdictions) has faced in attempting to ensure that only helpful and reliable scientific expert evidence is received in legal proceedings. Justice has been repeatedly threatened by the use of flawed expert evidence – in the Goudge Inquiry, it was pathology evidence, and in Broomfield it was toxicology evidence.

3. Original and Expanded Mandate

11. On November 26, 2014, the cabinet of the Government of Ontario established this Independent Review through Order in Council 1543/2014 (Appendix 1). In essence, the order in council directed a review of the scientific adequacy and reliability of MDTL’s hair-testing methodology and its use in criminal and child protection proceedings generally from 2005 to 2010.9 It also sought recommendations as to whether any additional review was warranted for individual cases or classes of cases.

12. My first actions as Independent Reviewer were to retain counsel and expert forensic toxicologists to advise on the science of hair testing. The Independent Review received a preliminary opinion from the expert scientists with respect to hair testing at MDTL between 2005 and 2010. This preliminary opinion explained that the immunoassay tests used by MDTL to assess drug use during this time were intended as screening tests only and were unreliable in the absence of another test (known as a confirmation test). The Independent Review’s experts concluded that the tests should not have been employed by MDTL to draw conclusions such as the ones reached in the Broomfield case.

13. The experts also raised concerns about whether MDTL’s laboratory procedures adhered to generally accepted international forensic standards. In particular, the experts requested that the Independent Review obtain additional information from MDTL and the Hospital regarding the Laboratory’s submissions to an international body that conducted proficiency testing.

14. After the Hospital considered the content of the preliminary opinion prepared by the experts and reviewed the Laboratory’s proficiency-testing submissions, the Hospital announced, on March 5, 2015, that it had decided to suspend all non-research activities of MDTL on an interim basis. On April 17, 2015, the Hospital announced its decision to close MDTL on a permanent basis. In light of the investigations that the Independent Review had conducted to that date and the preliminary opinion received from the experts, I met with the Attorney General. I recommended the expansion of the Independent Review’s mandate to include the review of the adequacy and reliability of the hair-testing methodology used by MDTL for alcohol as well as drug testing for the period expanded to March 5, 2015, when the Hospital suspended the Laboratory’s non-research operations.

15. On April 22, 2015, Order in Council 1543/2014 was revoked and replaced by Order in Council 449/2015, which expanded the mandate of the Independent Review to include

  • the period between 2005 and 2015;
  • an assessment of the extent to which MDTL’s operations met internationally recognized forensic standards;
  • an assessment of MDTL’s alcohol testing in hair; and
  • other matters related to the operation of MDTL that I considered necessary and appropriate to review.

16. The mandate also continued to require me to make recommendations on whether my findings warranted an additional review or process in respect of individual cases or classes of cases, and, if so, the nature and extent of any such review or process. The complete order in council is appended as Appendix 2.

17. My mandate includes addressing the manner in which MDTL’s hair test evidence was used in particular classes of legal proceedings, but restricts me from reporting “on any individual cases that are, have been or may be the subject of child protection or criminal investigations or proceedings.” In other words, my mandate was to look at systemic problems, largely of a scientific nature. I did not look at individual cases that may have been affected by flawed MDTL test results. By the terms of the order in council, any individual case review will be subject to the subsequent review or process that I recommend in this Report.

4. Setting Up the Review

4.1 Review Counsel

18. I was fortunate to retain Linda Rothstein as lead counsel to the Independent Review. Ms. Rothstein, a highly regarded counsel, has significant inquiry experience. Most relevant, Ms. Rothstein was commission counsel on the Goudge Inquiry. Her considerable experience and talent have been invaluable to the Independent Review. Her familiarity with the forensic landscape has enabled the Independent Review to proceed expeditiously and efficiently. Robert Centa and Tina Lie, both of whom worked with Ms. Rothstein on the Goudge Inquiry, were also central to the work of the Review. When the mandate expanded in April 2015, Jodi Martin joined the legal team and was a great help during the balance of the Review. This legal team worked diligently to ensure the Independent Review’s goal of approaching the investigation efficiently, objectively, and fairly. I am very grateful to the members of the legal team for their investigative work, their comprehensive research, and their legal advice on all aspects of this Independent Review.

4.2 Scientific Experts

19. I engaged two of the world’s leading experts in forensic toxicology to assist me with the Independent Review.

20. Dr. Gail Audrey Ann Cooper has worked in the field of forensic toxicology for more than 15 years, including for seven years within the academic unit of Forensic Medicine and Science at the University of Glasgow, where she was lead consultant forensic toxicologist. She is currently an independent consultant forensic toxicologist and director of Cooper Gold Forensic Consultancy Ltd., based in Scotland. Dr. Cooper completed her doctorate in forensic toxicology at the University of Glasgow in 1999, has published numerous peer-reviewed studies and book chapters, and is co-editor of a key textbook on forensic and analytical toxicology. Since 2010, she has been a member of the board of directors of the Society of Hair Testing (SoHT), the leading international organization in this subspecialty of forensic toxicology. In that capacity, she has contributed to the creation of the consensus statements published by the SoHT, which contain many of the relevant international standards on hair testing. She has recently accepted the post of director of forensic toxicology at the Office of the Chief Medical Examiner in New York City.

21. Professor Olaf H. Drummer is a forensic pharmacologist and toxicologist. He is the deputy director (academic programs) at the Victorian Institute of Forensic Medicine in Melbourne, Australia. He is also the head of the Department of Forensic Medicine at Monash University, also in Melbourne. Professor Drummer completed his doctorate in the faculty of medicine (pharmacology) from Melbourne University in 1980 and has been involved in the field of forensic toxicology for 40 years. He has published extensively in the fields of forensic pharmacology and toxicology. Professor Drummer was president of the International Association of Forensic Toxicologists (TIAFT) from 2008 to 2011. Notably, Professor Drummer was a contributor to both the 2001 and 2014 United Nations International Drug Control Programme’s Guidelines for Testing Drugs under International Control in Hair, Sweat and Saliva, a publication also widely recognized as containing many of the international standards relating to hair testing.

22. Dr. Cooper and Professor Drummer have been essential members of the Independent Review team. They educated me on the science of hair testing and international standards of forensic toxicology. They came to Toronto for several days at a time to ensure that I understood the underlying scientific principles and issues. Dr. Cooper visited Toronto in January 2015, shortly after the initiation of the Independent Review. Both Dr. Cooper and Professor Drummer came to Toronto in March 2015 and again in July 2015, after the expansion of my mandate to include the years 2010 to 2015 as well as hair testing for alcohol. During their stay in Toronto, they visited MDTL and were able to see first-hand how the Laboratory operated, albeit in its new setting. Throughout the Independent Review, Dr. Cooper and Professor Drummer have made themselves available for conference calls and endless communications to assist us in understanding and writing about the analytical and interpretive aspects of hair analysis and the international forensic standards by which MDTL’s operations are being measured. I am most grateful for their expertise and their dedication to the work of the Independent Review, both of which were critically important to the fulfillment of my mandate.

4.3 Administrative and Editorial Support

23. Joanna Arvanitis, my executive coordinator, was seconded from the Ministry of the Attorney General to provide administrative support to the Independent Review. Ms. Arvanitis was an invaluable member of the team. Among her many tasks and talents, she liaised with necessary contacts and supports, efficiently managed the submissions and documents that the Independent Review received (a particularly large and important responsibility), and ensured that the day-to-day operations ran smoothly. I appreciate the time that she took away from the ministry to assist us with the Review. I am also grateful for the expertise of our three editors – Rosemary Shipton, Mary McDougall Maude, and Dan Liebman – whose contribution to this Report was invaluable.

5. Process of the Review

24. This Independent Review is not a public inquiry under the Public Inquiries Act.10 A public inquiry will usually involve a public hearing, where oral evidence is tested by examination-in-chief and cross-examination. A commissioner conducting a public inquiry is usually granted specific powers in his or her mandate, including the power to compel productions of documents and attendance at the hearing.

25. A review is quite different. Generally, a review is a focused investigation into a narrow and well-defined issue – in this case, the adequacy and reliability of particular scientific tests for forensic purposes. A review is designed to give advice to the government regarding the issue of concern. It usually does not involve a public hearing. In addition, unlike in a public inquiry, a reviewer does not have powers of compulsion. In other words, in carrying out my mandate, I could not compel individuals or institutions to produce documents, nor could I compel individuals to meet with my team or answer our questions. In the case of this Review, the order in council specified the Hospital’s acknowledgment of the need for the Independent Review. The order in council also specified that all Ontario ministries would assist the Review “to the fullest extent possible.”

26. The order in council created few procedural requirements. For example, it permitted but did not require “public and/or private meetings and interviews.” Accordingly, I was given the flexibility to develop a process that would best address my core mandate. The process that was implemented enabled me to carry out a private but thorough, efficient, cost-effective, and fair investigation, which, at the same time, allowed the major participants to have input on the matters relevant to my mandate. The Independent Review gathered information in six ways: review of primary documents, in-person and written interviews with current and former MDTL staff, consultations with organizations, requests for and review of submissions filed with the Independent Review, examination of reported court decisions, and roundtable meetings to discuss the impact of the Laboratory’s hair test results and appropriate recommendations.

5.1 Primary Document Review

27. Given the nature of my mandate, it was apparent that I would rely heavily on documents and information within the control of the Hospital. Early on, with input from the Review’s experts, my counsel worked with the Hospital to identify categories of documents for production. As the Independent Review continued, the Hospital responded to further requests arising from the document review and interviews.

28. The Hospital provided the Independent Review with access to many of MDTL’s paper and electronic records. The Independent Review examined thousands of documents in total. In particular, in order to assess the adequacy and reliability of the Laboratory’s hair-testing methods and procedures, the Independent Review randomly selected and reviewed sample case files for each of the drugs (as well as for alcohol markers) tested in each of the years between 2005 and 2015. The random case files reviewed included the raw data that were used to generate the Laboratory’s results, to the extent that the Hospital had those records. The Independent Review did not ask the Hospital to produce all relevant documents, but rather made targeted document requests based on the information obtained and the advice of our experts.

29. In addition to providing access to the documents under its control, the Hospital prepared a Preliminary Institutional Report as well as an Oversight Summary. The former provided an overview of the Hospital’s operations at an institutional level, and the latter described the extent of the Hospital’s oversight of MDTL during the 10-year period under review.11

5.2 Interviews with Current and Former MDTL Staff

30. The most practical means of obtaining information in a timely and efficient manner was to interview the key individuals involved. Given the nature of the Independent Review, individual participation was voluntary. The interview process was designed to be as fair and respectful as possible to all involved and to gather evidence in an independent and impartial way. For example, after an interview, my counsel prepared interview summaries and gave them to the person interviewed to ensure the accuracy of the summary. Throughout this Report, when I quote the words of an individual who has been interviewed, I am quoting from the interview summaries that my counsel prepared and the individual approved.

31. Review counsel conducted extensive interviews with a number of individuals, including Joey Gareri. Mr. Gareri, who was MDTL’s laboratory manager from May 2005, was the lead in the Hospital’s search for documents and for answers to many detailed and numerous questions. He has worked diligently over the past several months to find answers for the Independent Review’s ongoing questions.

32. Dr. Gideon Koren, MDTL’s laboratory director throughout the period of the Independent Review, and Julia Klein, MDTL’s laboratory manager until April 2005, both declined my counsel’s requests for an interview. However, through their counsel, they provided written submissions and answers to written questions (and follow-up questions) from my counsel. When I refer to what Dr. Koren and Ms. Klein told the Independent Review throughout this Report, I refer to the written submissions and answers that they provided.

33. In addition, the Hospital facilitated a presentation for the Independent Review. As well, the Review team visited the Laboratory for a tour (recognizing that MDTL moved into its current location only in 2013). In addition, the Hospital gave my counsel and experts access to other individuals who worked at MDTL during the relevant period and who could answer questions regarding the Laboratory’s testing procedures.

5.3 Consultations with Organizations

34. The Independent Review also met with a number of organizations that provided us with valuable background, context, and information. My counsel or I met with representatives from the Ontario Association of Children’s Aid Societies, child protection agencies, the Family Lawyers Association, the Office of the Children’s Lawyer, Legal Aid Ontario, the Office of the Provincial Advocate for Children and Youth, the Institute for Quality Management in Healthcare, and the toxicology division of the Centre of Forensic Sciences.

35. In addition to meeting with us, the Provincial Advocate for Children and Youth arranged for me to meet with youth who had experience with Ontario’s child protection framework. I am grateful to the seven young people who participated in that discussion.

36. As contemplated by the order in council, the Independent Review received full co-­operation from all ministries and all agencies, boards, and commissions of the Government of Ontario that my counsel contacted. In particular, the Independent Review met with representatives from the Ministry of Child and Youth Services, which oversees child protection agencies, and the Ministry of Health and Long-Term Care, which oversees hospitals and clinical laboratory accreditation. The assistance we received from these bodies and individuals was helpful and essential for the timely discharge of my mandate.12

5.4 Submissions

37. In March 2015, I issued a call for submissions from the public on any matters within the original mandate, and in June 2015 I issued a further call for submissions on matters within my expanded mandate. In addition, I invited organizations, including child protection agencies and legal organizations representing the family and criminal law bar, to provide submissions to the Independent Review on matters within my mandate.

38. In response to the March call, the Independent Review received more than 45 submissions from individuals and groups. In response to the June call, the Review received more than 35 submissions from individuals and groups. A list of the submissions we received is included as Appendix 4 to this Report.

5.5 Examination of Court Decisions

39. Although my mandate does not extend to reporting on individual cases, I reviewed reported decisions from the Ontario Court of Justice and the Superior Court of Justice in which MDTL results or evidence was relied on by a child protection agency or a criminal court.13 My review of these decisions provided helpful information about when, how, and why MDTL test results were presented to the court, and the extent to which parents, child protection workers, lawyers for the parties, and judges relied on those results in child protection or criminal proceedings. The cases reviewed are listed in Appendix 5.

40. Not all decisions are reported, and many decisions are given orally (oral decisions are usually not reported). Accordingly, the decisions researched did not include all the cases in which MDTL hair test results were tendered as evidence, nor was it necessary to include all decisions for the discharge of my mandate.

5.6 Legal Experts and Roundtables

41. As will be seen in Chapter 11, I have recommended that the Government of Ontario appoint a commissioner to conduct an individual review of certain cases or classes of cases. To assist me in considering the scope and mandate of a Second Review, I convened two roundtables of experts in child protection and family law. In advance of the roundtables, I retained Professor Nicholas Bala of Queen’s University Law School and Professor Shelley Kierstead of Osgoode Hall Law School to ensure I appreciated the operation of the Child and Family Services Act and the range of remedies that would be statutorily available or precluded in the event of a further review.14

42. I then convened an initial roundtable with the two academics and Lorne Glass, a highly regarded child protection lawyer who has acted for Jewish Family and Child Service of Greater Toronto as well as for parents and children involved in child protection. The initial roundtable discussed the recommendations that I was considering. This discussion advanced my thinking.

43. The second family law roundtable built on the work of the first. Along with the participants from the first session, the following individuals attended at the instance of their respective organizations or in their own capacity: the Honourable Justice Heather Katarynych of the Ontario Court of Justice; Bruce Rivers, executive director of Covenant House (who has long-term and varied child protection experience); Kristina Reitmeier (counsel with the Children’s Aid Society of Toronto); and Bill Sullivan (a highly regarded child protection lawyer for parents and children, who was recently appointed to the Ontario Court of Justice). Their comments at this roundtable improved significantly the recommendations I make in Chapter 11.

44. I also received valuable advice from Mark Sandler, one of Ontario’s leading criminal lawyers. Mr. Sandler and I discussed issues related to the criminal proceedings that fell within my mandate and the recommendations that I was considering.

45. These broad consultations were helpful, and I wish to thank all participants in both roundtables for their dedication, enthusiasm, and commitment to child protection in the province and their advice to the Independent Review. Any merit that accrues to the recommendations contained in Chapter 11 belongs to those giving me the wisdom of their experience. Any flaws are my own.

6. Nature of Findings

46. Consistent with the terms of the order in council, in this Report I do not express any conclusion or recommendation regarding professional discipline matters involving any person or the civil or criminal liability of any person or organization. Nothing in this Report should be interpreted as suggesting otherwise.

47. I was able to make the findings of fact necessary to discharge my mandate on the basis of the documents we reviewed, supplemented by largely uncontroversial witness interviews. In the result, my conclusions relate to, and should be read as, systemic rather than individual findings.


Notes

5 R v Broomfield, 2014 ONCA 725 (Broomfield).

6 After the child’s recovery, albeit with long-term health consequences, the child eventually moved in with relatives.

7 MDTL’s laboratory manager testified to the same effect at the 2007 preliminary inquiry.

8 Ontario, Inquiry into Pediatric Forensic Pathology in Ontario, Report, 4v (Toronto: Ministry of the Attorney General, 2008) (Commissioner Stephen T. Goudge).

9 The original mandate covered the years from 2005 to 2010 because MDTL had advised that it changed its testing methodology in 2010.

10 Public Inquiries Act, 2009, SO 2009, c 33, Schedule 6.

11 The Hospital prepared the Oversight Report based on its interviews of key persons and the review of relevant documents.

12 See Appendix 3 for a list of organizations and roundtables.

13 A list of child protection cases we reviewed is found in Appendix 5.

14 Child and Family Services Act, RSO 1990, c C.11.

Chapter 3
Introduction to Hair Analysis

1. Introduction

1. Hair testing to detect drug or alcohol use is a relatively new scientific field that has gained increasing prominence over the past two decades. The Society of Hair Testing (SoHT), the only international body dedicated to the development and standardization of hair-testing methodologies, was founded in 1995 and released its first consensus statement, which contained recommended guidelines on the examination of drugs in hair, in 1997.

2. One of the benefits of hair tests over other methods of drug or alcohol testing is the potential to provide insight into drug or alcohol use over time. However, hair testing is not without its challenges. Many factors – some that are within the hair-testing laboratory’s control and many that are not – affect the conclusions that may be drawn from a particular result. In addition, there are limits to what hair testing can say in a specific case. Hair tests alone cannot, for example, provide complete answers about whether a person has used drugs or alcohol – or when, how frequently, in what quantity, or the effect of that use.

3. Before I assess the adequacy and reliability of the hair tests performed by the Motherisk Drug Testing Laboratory (MDTL or the Laboratory), it is important for readers to understand generally the science of hair tests, including how they ought to be carried out, as well as the various factors that affect how the results should be interpreted.15

2. Hair Testing as a Method of Drug or Alcohol Testing

4. Our hair contains remnants of the substances that we consume or to which we have been exposed. Because hair testing involves the detection of compounds along the length of a hair sample, it has several advantages over urine or blood testing: Urine and blood testing can detect drugs or alcohol only in the few hours or perhaps the few days before the test. As a result of the small window of detection, individuals who are required to submit to routine drug or alcohol testing may abstain from these substances before a scheduled urine or blood test in an attempt to secure a negative result. Hair testing cannot be manipulated in this manner, as intermittent abstinence will not affect the results of the test. In addition, a single hair sample can be used to test for several months of use or exposure at a time, depending on the length of the hair and the amount of the sample available.

5. There are also advantages in the collection and storage of hair samples. For most people (although perhaps not for those who have religious or cultural practices that restrict haircutting), the collection of head hair is less intrusive than urine and blood testing. The collection of head hair is also less awkward than the observed collection of urine. In addition, storage is easier because, unlike urine or blood, hair samples do not require refrigeration and can be stored indefinitely. Because drugs are relatively stable in hair, a second representative hair sample can also be collected and analyzed if needed (assuming the person’s hair has not been cut or chemically treated since the first collection).

2.1 Distinguishing between Use and Exposure for Drugs of Abuse

6. The precise biochemical mechanisms through which drugs incorporate into hair are not fully understood at this time. Hair has different affinities and binding capacities for different drugs, and the manner in which a drug binds with hair may be unique to the particular drug. For example, basic drugs (such as cocaine, opiates, and amphetamines) have been found to incorporate into hair to a greater extent than neutral or acidic drugs (such as benzodiazepines and cannabis).16 The result is that the former drugs are typically present in higher concentrations than are the latter.

7. Although the precise mechanisms may not be known, it is widely accepted that there are three routes of incorporation of drugs into hair. Figure 3.1 depicts these routes:

  1. Drugs can diffuse from the blood supplying the hair follicle during the formation of the hair in the scalp.
  2. Drugs can incorporate into hair through sebum (which is an oily secretion from a gland under the surface of the skin called the sebaceous gland) and sweat. As the hair follicle grows, it bathes in sebum for several days before it breaks through the surface of the skin. Sweat from the sweat gland can also come into contact with the hair at the surface of the skin.17 When the hair is exposed to sebum and sweat, drugs that are contained in those secretions can diffuse into the hair.
  3. Because of the large surface-to-volume ratio of hair, drugs can easily incorporate into hair through contamination from the environment (referred to as “external contamination” or “passive exposure”). External contamination can occur in a variety of ways, including as a result of direct physical contact with a drug (e.g., if a person handles a drug and then touches her or his hair or that of a child) and exposure to the drug in the air (e.g., from dust or smoke).

8. The first two of these routes of incorporation are generally considered to result from active drug use (or ingestion) – which, depending on the drug, may include smoking, snorting, oral ingestion, or intravenous injection – because the drugs that incorporate into the hair originate from the body (either from the bloodstream supplying the hair follicle or from the sweat or sebaceous glands). The third route of incorporation does not result from active use of the drug, but rather from passive exposure to the drug in the environment.

9. When a hair test detects drugs in a hair sample, there is no way of determining how the drug was incorporated into the sample – whether it was through the body as a result of active drug use or through external contamination. A person who has not used drugs but who has been exposed to an environment that contains them may therefore test positive for the drug.

Figure 3.1 Incorporation routes of drugs into the hair. Source: R. Kornstrand, K. Scott (see Appendix 6). Reprinted with permission.

Figure 3.1 Incorporation routes of drugs into the hair
Source: R. Kornstrand, K. Scott (see Appendix 6). Reprinted with permission.

10. As a result, a hair test on its own cannot establish definitively whether a person used drugs or was simply exposed to them in her or his environment. For this reason, and as discussed below, proper washing procedures designed to remove or reduce external contamination are crucial if the hair test result is being interpreted to indicate use of a drug rather than just exposure to it. However, even with the best decontamination procedures, the possibility of external contamination cannot be ruled out entirely.

11. That said, the detection of metabolites can assist in distinguishing between active use and passive exposure when interpreting a particular test result. Metabolites are compounds that are formed when the body breaks down a drug during the metabolism process. Since they are created inside the body, the detection of metabolites is one of the tools that can assist in distinguishing between use (ingestion) of and external exposure (contamination) to the drug. However, some metabolites have been found to exist outside the body as well – as a result, the detection of metabolites, on its own, is not necessarily enough to suggest that a person used, rather than was exposed to, a drug. The primary metabolites of cocaine are benzoylecgonine, norcocaine, and cocaethylene;18 and the primary metabolite for heroin is 6-AM.

2.2 Bulk Analysis vs. Segmental Analysis for Drugs of Abuse

12. A hair sample may be tested for drugs of abuse in two ways: (1) the length of the hair may be tested as a whole (a process referred to as bulk analysis); or (2) the sample may be cut and tested in smaller segments (which is referred to as segmental analysis). Because hair grows an average of 1 cm per month, when a hair sample is tested using segmental analysis, the sample is typically cut into segments measuring between 1 and 3 cm, representing approximately one to three months of average growth per segment.

13. The advantage of segmental analysis is that it can provide insight into drug use or exposure within a more defined period than bulk analysis. This benefit is illustrated, for example, in the hair strand depicted in Figure 3.2 – where, in a 12 cm length of hair, the drug is contained only in the 1–2 cm section. If the entire 12 cm length was tested in bulk, a positive result would indicate only that there was use or exposure in the previous 12 months. By contrast, if the hair sample was cut into and tested in four 3 cm segments, the results may provide a more detailed profile; the results of the 0–3 cm segment closest to the scalp would approximately represent the most recent three months before the sample was collected, the 3–6 cm segment would approximately represent the three months before that, and so on. Testing the hair sample segmentally would thus provide a more precise history by indicating that the use or exposure occurred in the approximately three months before sample collection.

Figure 3.2 Example of 12 cm length of hair and segmentation

Figure 3.2 Example of 12 cm length of hair and segmentation

14. Segmental analysis has the added benefit of revealing any changes in drug concentration over time, as a lower concentration in the segment closest to the scalp may suggest the individual has used or been exposed to a lower amount of drugs in the most recent months than in the more distant past (and vice versa).

2.3 Testing for Alcohol Markers in Hair

15. Testing for ethanol in blood or urine can be used to determine blood alcohol content in order to assess a person’s level of alcohol use at a particular time. However, because alcohol is eliminated from the body within hours of consumption, blood or urine tests cannot assess the degree of alcohol consumption over an extended time frame. Hair testing, on the other hand, has the potential to do so.

16. Unlike blood and urine tests, hair analysis for the purposes of assessing alcohol use does not test for alcohol (ethanol) directly. Instead, hair tests have been developed to test for substances called “alcohol markers,” which are compounds (metabolites) produced by the body following the consumption of alcohol. Alcohol marker testing is a relatively recent development in the field of hair testing.19

17. The main alcohol markers currently tested in hair are fatty acid ethyl esters (FAEEs) and ethyl glucuronide (EtG). FAEEs comprise a group of more than 20 minor metabolites of ethanol. The SoHT recommends using the sum of four specific FAEEs: ethyl myristate, ethyl palmitate, ethyl oleate, and ethyl stearate. FAEEs are primarily incorporated into hair through sebum, which bathes the hair beneath the surface of the skin for several days before it grows out of the scalp. Because sebum or its residue can remain on hair and may not be washed out with regular shampooing, the longer the hair comes into contact with sebum or its residue, the more likely FAEEs will incorporate into the hair. The result is that concentrations of FAEEs tend to increase from the root to the far end of the hair.

18. The other alcohol marker that is tested in hair, EtG, is also a minor metabolite of ethanol. EtG is incorporated into hair predominantly through the bloodstream and, to a lesser extent, from sebum or sweat.

19. Hair samples are not segmented for alcohol marker testing. As a result of the manner in which FAEEs incorporate into hair (primarily through sebum), the SoHT recommends using defined hair lengths to test for FAEEs – that is, only the first 3 cm or 6 cm of the hair (measured from the scalp). Although sebum plays a smaller role in the incorporation of EtG into hair, the SoHT also suggests using defined hair lengths of 3 cm or 6 cm for EtG testing. Accordingly, alcohol marker testing can be used to assess alcohol consumption only in approximately the three or six months before the sample was collected, depending on the length of the sample being tested.

3. The Hair-Testing Process

20. In this section, I provide an overview of the hair-testing process generally, from the initial collection of the sample to the reporting of the test result. The process described below is not a description of MDTL’s practices during the relevant period, but rather is intended to provide the background needed to assess MDTL’s analytical practices.

3.1 Sample Collection

21. One of the benefits of hair testing is that samples do not need to be collected in specialized facilities and the collector need not have special qualifications. However, to minimize the risk of contamination and ensure that best practices are followed, the collector should be trained in sample collection.

22. Because hair testing is used to assess historical use or exposure to drugs or alcohol over the length of the hair, it is important to cut the hair as close to the scalp as possible. The SoHT recommends using hair from the posterior vertex region of the head, which has the least variation in hair growth (see Figure 3.3). Collectors should obtain enough hair to carry out the test and any follow-up if required. The amount is typically described as a “lock of hair” or a “pencil thickness of hair.”20 A laboratory establishes the minimum sample volume needed for a test (which depends on the sensitivity of the testing instrument). In the case of segmental analysis, that minimum sample volume would be needed for each segment, and the smaller the segment, the more strands of hair are needed to make up the required sample volume. As a result, collecting a sufficient volume of hair is particularly important when carrying out segmental analysis.

Figure 3.3 Posterior vertex region of the head

Figure 3.3 Posterior vertex region of the head

23. Hair collectors are typically given a hair collection kit that includes a chain-of-custody or test request form, along with instructions and the material required to collect the sample: foil, a collection envelope, security seals, an evidence bag, and a transportation envelope. The accepted best practice is to wrap the hair sample inside the foil, identify which end of the sample is the root end, and then send the sealed sample to the laboratory for testing.

24. The chain-of-custody form is used to document the date and time of collection as well as other information about the sample and its donor, including hair length, colour, and condition, as well as declared cosmetic treatments and prescribed medications. Because chain of custody is an important principle in forensic science, the form also acts as a record of those who have handled the hair sample from collection until receipt by the forensic laboratory.

3.2 Sample Preparation and Analysis

25. Once the hair sample is received by the laboratory, it must be prepared for analysis before it is tested. Individual laboratories have their own procedures, but Figure 3.4 illustrates the key stages in the preparation and analysis of hair samples for drugs of abuse.

Figure 3.4 Key stages of sample preparation and analysis of hair samples for drugs of abuse

Figure 3.4 Key stages of sample preparation and
analysis of hair samples for drugs of abuse

26. A description of each of the stages of analysis is provided below. As the testing process for alcohol markers differs in some respects from the process for drugs of abuse, those differences are also described below.

3.2.1 Sample Preparation
3.2.1.1 Segmentation

27. The first step in the sample preparation process for testing for drugs of abuse is to segment the hair (assuming segmental analysis is being carried out). Although the laboratory may receive information from the customer about whether segmental analysis is being requested and over which periods, in some cases the volume of the hair collected or the length of the hair may make segmental or bulk analysis inappropriate, regardless of the customer’s request. If a sufficient amount of hair has not been collected to permit segmental analysis, for example, the laboratory may determine that bulk analysis is more appropriate (in which case the laboratory should communicate this information to the customer before testing). In addition, it is generally inappropriate to test long strands of hair in bulk, since there is a “dilution effect” as the longer sample – and greater volume of hair – dilutes the amount of drugs that are present and capable of being detected by the hair-testing instrument.21

28. As noted, segmental analysis is not recommended for alcohol marker testing. Instead, the laboratory cuts a fixed 3 cm or 6 cm length from the root end of the hair for the purposes of carrying out its FAEEs and/or EtG testing.

3.2.1.2 Washing

29. Washing the hair sample before testing is standard practice across hair-testing laboratories. The washing (or decontamination) step is important because it not only cleans visible dirt and grime from the hair that may interfere with the test but also reduces or removes contaminants that may not be visible, such as shampoo, hair gel, hairspray, and, importantly, drugs on the surface of the hair. In addition, washing can remove sebum and sweat on the hair, as well as residue from those secretions.

30. Although no single washing technique is advocated for any particular hair-testing method, it is widely accepted that washing or decontamination is required if the test is being administered to assess drug use. The SoHT has always recommended washing hair with a combination of water (or an aqueous-based dilute buffer) and an organic solvent.

31. Because there is no standardized approach to washing hair samples before testing, laboratories are expected to evaluate the effectiveness of their individual wash procedures. Ideally, wash procedures will be selected that remove, to the extent possible, all traces of surface contamination, without removing the drugs that have incorporated into the hair shaft or changing the profile of the drug present in the hair.

32. Although washing hair samples prior to analysis is recommended, it is not always mandatory before screening tests, given the purpose of those tests. In high-volume screenings where the results are predominantly negative (e.g., workplace drug testing), it is acceptable to carry out the initial screening test on unwashed samples to distinguish quickly between negative results and preliminary positive results. However, any time a hair sample is being tested to assess drug use rather than exposure, it must be washed before confirmation testing.22

33. The only reason a laboratory may decide not to wash a hair sample before carrying out the confirmation test is to detect the presence of the drugs or metabolites without regard to how they got into the hair (i.e., whether it was through active use of the drug or through external contamination). In such circumstances, however, the fact that the sample was not washed and the rationale for not doing so should be clearly set out in the results report.

34. In addition, the amount of drugs contained in the solution and solvent used to wash the sample (called the “washings”) can help in assessing the extent of external contamination. A significant amount of the drug in the washings relative to what was found in the hair would indicate that there was substantial external contamination. Moreover, knowing when the individual’s hair was last washed may provide some indication of the timing of exposure. For example, detecting a significant amount of drugs in the washings may suggest exposure in the past two days if the hair was washed two days ago. For these reasons, the SoHT has recommended that laboratories store the washings for further analysis, if necessary.

35. Finally, because FAEEs incorporate into hair primarily through sebum, it is particularly important for the laboratory to wash the hair before analysis for FAEEs. This washing step removes the secretions coating the surface of the hair and, to the extent possible, ensures that the test measures the amount of FAEEs that have diffused into the hair, rather than the amount of FAEEs that are contained in the sebum on the hair. Although contamination from sweat or sebum is less of an issue when testing for EtG, washing the hair is still important to remove any impurities that may interfere with the analysis.

3.2.1.3 Homogenization

36. After the sample is washed and dried, the laboratory either cuts the hair into small snippets of 1 to 2 mm using a pair of sharp-tipped scissors (a process referred to as “chopping”) or mills the hair into a powder. The purpose behind this step is to improve the homogeneity of the sample and to increase the hair’s surface area to enhance the potential to detect drugs or metabolites in the hair. Although powdering is widely regarded as the more effective means of homogenizing the sample, chopping into very small snippets is also an accepted practice.23

3.2.2 Sample Pre-treatment and Extraction

37. The next step in the sample preparation process is to soak (incubate) the cut or powdered hair in a solution that either extracts the drugs and metabolites (and alcohol markers) into the solution without damaging the hair or results in the complete digestion of the hair. In either case, the compounds are liberated from the hair into the solution, and it is the solution that is ultimately tested.

38. As with the washing procedure, there is no standardized approach to extraction; however, laboratories are expected to investigate the effectiveness of their extraction procedures. Because studies have shown that the chemical solutions used to extract the drugs from the hair may cause hydrolysis of certain drugs (e.g., cocaine and heroin), leading to the spontaneous creation of the drug’s metabolite, laboratories should have controls in place to optimize their extraction procedures while at the same time minimizing the risk of hydrolysis.

39. Because a laboratory can test only the amount of drugs or metabolites that it is able to extract from the hair, the extraction step is crucial to the sample preparation process. If a laboratory has poor extraction procedures, its instruments may not detect drugs that are actually present in the hair – resulting in false negative or falsely low results.

40. It is also important for laboratories to have standard extraction procedures, particularly when comparing test results with one another. For example, if a laboratory was to compare two sets of results for the same person, the first from a test carried out today and the second from a test carried out six months ago, an “apples-to-apples” comparison would be possible only if both tests followed the same extraction procedures. Put another way, if the laboratory made significant changes to its extraction procedures during the previous six months, it is possible that any difference in the results from six months ago and those from today was caused by the change in laboratory conditions rather than a change in drug use or exposure. Understanding the extraction technique followed for each test performed is therefore critical for interpreting the laboratory’s test results.

41. In addition, depending on the drug or metabolite being tested, a “cleanup” step may be required before carrying out any confirmation tests. Most drugs and compounds are extracted by using liquid-liquid extraction (LLE) or solid-phase extraction (SPE) in this step. LLE involves extraction of a substance from one liquid into another liquid phase, whereas SPE is another technique that involves separating out the substances in a mixture according to their physical and chemical properties. Headspace–solid-phase micro-extraction (HS-SPME), which uses a solid fibre coated with a polymer to attract the compounds within a sample, is another extraction technique that may be used for volatile drugs and compounds.

3.2.3 Testing for Drugs of Abuse

42. Drugs of abuse may be tested for in hair in two stages: a screening test (usually an immunoassay-based test) and a confirmation test. If a sample tests negative following the screening test, then, in most cases, no further tests are conducted. In contrast, positive results from an immunoassay-based screening test must undergo what is referred to in the science as a “confirmation test.” Only once the confirmation test has returned a positive result can a laboratory conclude that the sample contained the drug or metabolite and report the resulting concentration.

3.2.3.1 Immunoassay-Based Screening Tests

43. Immunoassay is a biochemical test that detects the presence of the target compound through the use of antibodies. Radio-immunoassay (RIA) tests were initially used to screen hair samples for drugs of abuse. However, over the past decade, if not longer, most hair-testing laboratories have replaced RIA with enzyme-linked immunosorbent assay (ELISA) tests. ELISA is what MDTL used from 2005 to 2015.

44. ELISA tests are widely used to screen for drugs of abuse. The ELISA test is sensitive and, when appropriately validated, provides a tool for quickly distinguishing between negative samples that do not require any further testing and positive samples that do. The tests are commercially manufactured and can be purchased as “kits” from a variety of suppliers. There are separate ELISA kits for each drug or class of drug. However, there are currently no commercially available ELISA kits for alcohol markers.

Figure 3.5 Image of 96-well ELISA plate (after addition of substrate but before end of incubation period)

Figure 3.5 Image of 96-well ELISA plate (after addition of substrate but before end of incubation period)

45. ELISA kits typically come in the form of a plate containing 96 wells, and each well comes coated with specific antibodies that have binding sites for the target drug. The ELISA test works by adding both the sample being tested and a solution containing an enzyme-linked version of the target drug to one of the wells. Once added, any drugs in the sample will compete with the enzyme-linked drugs for the binding sites on the antibodies in the well. A substrate (a compound that reacts with the enzymes that are linked to the drugs) is subsequently added, causing a colour change in the enzyme-linked drugs that have bound to the antibodies – from clear to blue. To stop the assay after a predefined time (called the incubation period), an acid is added, which changes any blue colouring to yellow.24 Figures 3.5 and 3.6 are images of a typical 96-well ELISA plate before and after the incubation period.

Figure 3.6 Image of 96-well ELISA plate (after incubation period and addition of acid

Figure 3.6 Image of 96-well ELISA plate (after incubation period and addition of acid)

46. The test works by measuring the extent of the colour change (called the optical density or absorbance) after the incubation period: The lower the amount of drugs present in the sample, the more intense the colour change because more enzyme-linked drugs will have bound to the antibodies (having had less competition for the binding sites). Conversely, the more drugs in the sample, the less intense the colour change because more of the drugs in the sample would have competed for and attached to the binding sites. The optical density is measured to assess if the result falls above or below a defined cut-off: If the former, it is considered positive, requiring further testing; if the latter, it is negative, and no further testing is required.

47. Because each ELISA kit comes with 96 wells, the ELISA test can be performed in batches, meaning that the test can be carried out simultaneously on multiple specimens. ELISA tests are therefore an efficient method to screen for drugs of abuse.

48. However, the antibodies in the wells can also bind with other compounds that are structurally related to the target drug, a phenomenon known as “cross-reactivity.” When the antibodies cross-react with structurally related compounds, the test may result in little or no colour change, even in the absence of the target drug or metabolite; in other words, a false positive.

49. Ultimately, the ELISA test is limited in what it can say about a particular sample because it relies on biochemical reactions: It does not separate out the target drug from the other substances contained in the sample. Nor does it identify the target drug in the sample. As a result, although ELISA can be used to measure the extent of the colour change in the solution, it cannot be used to measure the amount of the target drug contained in the sample. For these reasons, ELISA is not used to quantify drugs, but rather to distinguish quickly between samples that do not warrant further testing and those that do.

3.2.3.2 Confirmation Tests

50. Unlike the ELISA test, confirmation tests separate out the different compounds within a sample from one another using a technique called “chromatography.” Then, using a technique called “mass spectrometry,” they identify the separated compounds according to their mass spectral properties.

51. The confirmation tests relevant to the Independent Review are gas chromatography–mass spectrometry (GC-MS) and liquid chromatography–tandem mass spectrometry (LC-MS/MS). The principles underlying both techniques are similar. The sample solution is transported through a hollow column using a gas or liquid. The inside of the column is coated or packed with a chemical material. Because the different compounds within the solution travel through the column at different rates, they can be separated from one another as they interact with the chemical coating or the material inside the column. The process of separation creates what is called a “chromatogram” of peaks and valleys, much the same way that colours of ink can be separated on blotting paper. The size of the peaks can be used to quantify the amount of the separated compounds (the larger the peak, the greater amount of the compound). At the end of the column, the mass spectrometer fragments the separated substances into ions that are further separated according to their molecular weight (and ionic charge) and their relative intensity. The spectral properties of the fragmented ions are then used to identify the substances contained in the sample (akin to a molecular fingerprint).

52. In LC-MS/MS tests, tandem mass spectrometry (MS/MS) is the detection technique used at the end of the chromatographic process. When analyzed by MS/MS, the fragmented ions are further fragmented to form secondary ions, referred to as “daughter ions” or “transitions.” The spectral properties of the transitions are then analyzed to identify the substances contained in the sample.

53. Because of the manner in which the tests operate, both GC-MS and LC-MS/MS are widely accepted as reliable analytical methods for identifying as well as quantifying the amount of the target compounds present. GC-MS and LC-MS/MS have been called the gold standards for hair testing for drugs of abuse.

54. Another advantage of GC-MS and LC-MS/MS analysis is that, unlike ELISA, the instrument can be used to test for multiple drugs or metabolites at the same time. As a result, some laboratories use GC-MS and LC-MS/MS rather than ELISA to screen for drugs of abuse. Although these techniques are more expensive and less practical for screening large numbers of samples (because they cannot, for example, screen up to 96 samples at a time), they have the advantage of providing simultaneous screening and identification of many drugs in a single analysis.25 Even when chromatographic methods are used as a screening test, however, the recommended best practice is that results should still be confirmed by a second test before being reported as positive. Ideally, the confirmation test will use a different method from the screening test (e.g., use of GC-MS to screen and then LC-MS/MS to confirm).

3.2.3.3 A Note on Terminology

55. I note that forensic toxicologists often refer to positive results from an immunoassay-based test such as ELISA as a “presumptive positive.” Other terms that are used include “unconfirmed positive” or “tentative positive.”

56. However, it is important to be clear about the terminology used to describe results from an ELISA screening test as opposed to those from a confirmation test. Forensic toxicology requires documentation and terminology that serve the needs of the participants in the justice system. Because any misunderstanding of the meaning of the results could have severe consequences, information provided by forensic toxicologists must be communicated to the justice system in a way that is accurate, fully comprehensible, and unambiguous.

57. Although, to forensic toxicologists, it is implied and understood that a “presumptive positive” test is preliminary only and must be followed by a confirmation test, I do not think the phrase is sufficiently clear and unambiguous for use in the justice system or by this Independent Review. The phrase “presumptive positive” suggests that the result is positive or at least probably positive unless it has been disproved by a confirmation test. However, such a suggestion would be wrong; the ELISA test produces a preliminary result, not a presumptive one. As a result, for clarity, I will use “preliminary,” “unconfirmed,” or “tentative” to refer to ELISA test results that have not been confirmed.

58. Similarly, the term “confirmation test” implies that the test is designed to confirm the presence of a drug or metabolite that may be present. A confirmation test does not “confirm” anything; it determines at first instance if a drug or metabolite is present and, if so, the concentration. Put another way, without a confirmation test there is no identification of the drug in the hair. However, because the science refers to these tests as confirmation tests, I have continued to use that term in this Report.

3.2.4 Testing for Alcohol Markers

59. Unlike for drugs of abuse, no immunoassay-based screening tests for alcohol marker testing in hair are commercially available. Hair analysis for FAEEs routinely involves HS-SPME (described above) to extract the FAEEs from the sample before testing using GC-MS or LC-MS/MS. EtG is tested using GC-MS or LC-MS/MS following a process similar to the one used for drugs.

3.3 Analytical Concepts

60. Several other analytical concepts need to be understood in assessing the adequacy and reliability of a laboratory’s analytical methods.

3.3.1 Calibration

61. All analytical instruments need to be calibrated. Calibration involves the testing of several samples that contain known quantities of the target compound (referred to as “calibrators”). The results from testing the calibrators are plotted on a graph, called a “calibration curve,” which is then used to calculate the amount of identified drug in the sample.

62. The number of calibrators required depends on whether the test is being run for qualitative or quantitative purposes. A qualitative test simply seeks to determine if the test result is positive or negative, while a quantitative test determines the concentration of the target compound present. If testing for qualitative purposes, only two calibrators are required – one blank calibrator to establish the standard for a negative result, and one calibrator with a known quantity of the target compound to establish the standard for a positive result. If testing for quantitative purposes, at least five calibrators are usually needed – from low to high concentrations – in order to plot a sufficiently reliable calibration curve.

3.3.2 Quality Controls

63. Laboratories are required to use quality controls to assess if their tests have been run properly. This process typically involves running tests using a negative control (a blank sample that does not contain the target compound) and a positive control (a sample containing a known quantity of the target compound) with every batch. If the test has been run properly, the negative control should produce a negative result, and the positive control should produce a positive result close to the known concentration.

64. It is also common practice for laboratories to test more than one sample at a time, a practice referred to as “batch testing.” Monitoring batch-to-batch performance is important to ensure that the data generated are consistent with one another. Quality control performance across batches is assessed using a Levey-Jennings chart, which plots the dates of the test on the X-axis, and the results for the positive and negative controls on the Y-axis. The chart allows laboratories to identify and investigate statistically significant deviations that may be caused, for example, by errors in or other problems with the calibration.

3.3.3 Internal Standards

65. When carrying out GC-MS or LC-MS/MS analysis, laboratories use compounds called “internal standards” to correct for any potential differences in recovery arising from the sample preparation and the extraction process. Internal standards are drugs or metabolites that are almost identical to the target drug or metabolite. The most common (and recommended) ones are known as “deuterated internal standards” – their molecular makeup is identical to that of the target drug or metabolite, except the hydrogen ion has been replaced by deuterium (also called heavy hydrogen). For example, for cocaine, a deuterated internal standard referred to as “cocaine-d3” is used, in which three deuterium ions replace the hydrogen ions on the molecule.

66. Internal standards are added to every sample before it is tested using GC-MS or LC-MS/MS. The same concentration of the internal standard is used in every sample. After the test has been run, the laboratory calculates the concentration of the target drug in the sample by comparing the ratio of the drug to the internal standard with the same ratio in the calibrators. In this way, the internal standard corrects for any differences that may arise in the sample preparation and extraction process.

67. Because the GC-MS or LC-MS/MS instrument separates out the compounds in a sample, it can detect the internal standards separate and apart from the target drug or metabolite in the sample in order to carry out these calculations. By contrast, because ELISA responds not only to the target drug or metabolite but also to structurally related compounds (and ELISA tests may therefore cross-react with the internal standards for the target drug), internal standards are not used in tests carried out using ELISA.

3.3.4 Cut-offs

68. The way to determine if a test result is positive or negative is by assessing if the result falls above or below a certain cut-off. Two limits can affect the appropriate cut-off for a drug or metabolite:

  1. The limit of detection (LOD) is the lowest concentration of the substance that the instrument can reliably detect.
  2. The limit of quantification (LOQ) is the lowest concentration of the substance that the instrument can reliably quantify.

69. It is up to individual laboratories to determine, through instrument validation, the LOD and the LOQ for each of the tests they carry out.

70. The SoHT has recommended cut-offs for reporting positive results for drugs of abuse; however, those cut-offs are only guidelines, and it is ultimately up to the laboratory to develop its own cut-offs for assessing the results that it generates. For example, to eliminate small positive results that could have occurred through external contamination, some laboratories may choose to adopt a cut-off that is higher than the LOQ of their instrument. Other laboratories may adopt the LOQ as their cut-off and report the concentration of any sample that is higher than the LOQ.

71. For alcohol markers only, the SoHT has developed cut-offs on alcohol use. These cut-offs, which are unrelated to an instrument’s cut-off, are described below.

3.4 Quality Management System

72. In addition to running quality controls with each batch to ensure that the specific test has been properly carried out, laboratories should have a quality management system, which refers to the overarching processes that exist to ensure quality in the results that the laboratory reports. Core components of a quality management system include, at a minimum, documentation of the laboratory’s policies and procedures – among them, its analytical methods – in standard operating procedures. As part of the quality management system, laboratories should also have a system in place to identify and correct errors in the testing process.

4. Interpretation of Hair Test Results

73. After the laboratory has generated the hair test result (following the above processes), the next question is to determine what that result means. A number of factors affect drug concentrations in hair, complicating the interpretation that can be given to a particular result. In this section, I set out the main factors that may affect the interpretation of hair test results. Given these complications, interpretation of hair test results should be provided only by properly trained and qualified experts who have considered all the factors, scientific and non-scientific, that may have contributed to the results.

4.1 Hair Colour Bias

74. Natural hair pigmentation has been found to affect the incorporation of certain drugs into hair. Basic drugs, such as cocaine, opiates, and amphetamines, have been found to incorporate more readily into darker hair (which contains more of the pigment eumelanin) than in lighter-coloured hair (which contains less eumelanin). As a result, if two people used the same amount of cocaine, for example, the person with black hair would be expected to have a higher concentration of cocaine in her or his hair than the person with blond hair. In fact, as early as 2000, studies have shown that the difference can be substantial, with black-haired individuals showing up to 10 times the drug concentration as people with lighter-coloured hair.26 In contrast, hair colour has less of an effect on the incorporation of non-basic drugs, such as benzodiazepines and cannabis, into hair.

75. This difference among individuals with different hair pigmentation is a phenomenon referred to as “hair colour bias.” Given its potential impact on test results, hair colour bias must be taken into account whenever a comparison is drawn between test results from different individuals.

76. As the incorporation of FAEEs and EtG into hair is not affected by natural hair colour, hair colour bias is not an issue for FAEEs and EtG analyses.

4.2 Effect of Hair Products, Cosmetic Treatments, and Hair Damage

77. Because drugs and metabolites are not permanently bound to the structures in the hair, hair products, cosmetic treatments, and damage to the hair shaft can affect drug, metabolite, and alcohol marker concentrations in a hair sample.

78. Drug and metabolite concentrations in hair have been found to decrease over the length of the hair because of a natural “washing out” of the drugs and metabolites over time – particularly for the more water-soluble drugs, such as many opiates (including codeine, morphine, and 6-AM). In addition, sunlight and weathering have been shown to reduce concentrations of certain drugs in hair, including cocaine, benzoylecgonine, 6-AM, cannabis / THC, and methadone. Cosmetic treatments, such as dyeing, bleaching, and perming, can further reduce drug concentrations in hair. Despite these effects, research has shown that it is unlikely for regular washing and/or the use of aggressive cosmetic agents to remove all traces of a drug or metabolite in the hair.

79. In contrast, cosmetic products containing humectants, which are found in some specialized hair care products used to hydrate the hair, can have the opposite effect. Because humectants work by penetrating the hair, a recent study has shown that they may lead to an increased risk of false positives, as any drugs on the surface of the hair may penetrate the hair with the humectants.27 In addition, humectants may promote the binding of the drugs to the inside of the hair, making them more difficult to remove with normal washing procedures.

80. As well, damage to the hair shaft can make the hair more permeable to surface absorption from exposure to drugs, particularly for opiates and, to a lesser extent, cocaine. As a result, although damage to the hair from the weather or cosmetic agents can decrease the amount of drugs or metabolites in in the hair, the hair, as it becomes more damaged, can be more susceptible to external contamination.

81. As with drugs, cosmetic treatments such as dyeing, bleaching, or perming can decrease both FAEEs and EtG concentrations in hair. However, hair products with high ethanol content (e.g., certain hair sprays, gels, and hair lotions) have a particular effect on FAEEs tests by increasing (in some cases, substantially) the concentration of FAEEs in the hair. Significantly, the increase of FAEEs caused by ethanol-containing hair products is not an external contamination issue that can be mitigated through robust washing procedures; the use of ethanol-containing hair products causes FAEEs to incorporate into the hair, which cannot then be washed away. In recent years, the sensitivity of FAEEs to ethanol-containing products and treatments has caused many hair-testing experts to question the use of FAEEs as an alcohol marker. Because EtG results are unaffected by ethanol-containing hair products, EtG is now widely accepted as the more reliable alcohol marker in hair.

4.3 Interpreting Timing of Use or Exposure

4.3.1 Variation in Hair Growth Rates

82. Although most hair-testing laboratories assume an average growth rate of 1 cm per month for head hair, this assumed growth rate is just a generalization. A number of factors affect hair growth, including age, gender, pregnancy, metabolic and genetic disorders, nutrition, and even the season. The growth rate of head hair commonly ranges from approximately as little as 0.6 to as much as 1.5 cm per month. Other studies have reported growth rates ranging up to as much as 3.36 cm per month,28 with differences across ethnicities, genders, and ages.29

83. These varying growth rates can have a significant impact on the interpretation of a particular test. If an individual’s hair grows 0.6 cm per month, for example, then a 3 cm segment of hair would reflect approximately five months of use or exposure to drugs. Conversely, if the person’s hair grows at a rate of 1.5 cm per month, the same 3 cm segment would reflect only approximately two months of exposure.

84. In addition, growth rates in infants under one year are different from those of adults or children. The actual presence of hair varies among newborns, and, regardless of the amount of hair present at birth, newborn hair tends to fall out at approximately three months of age. It is only by approximately six months that the hair that emerges begins to resemble that of an older child. As a result, the average growth rates used in testing adult hair generally do not apply to infants under 12 months, and it can be difficult to identify a window of detection when testing for drugs in infants and young children.

4.3.2 Banding

85. Even putting aside the varying growth rates among individuals, head hair within an individual does not grow at a constant rate, but rather in a series of phases:

  1. The anagen phase, which lasts two to six years, is the growing phase of the hair. It represents approximately 85 percent of head hair (particularly hair taken at the back of the head) at any given time.
  2. The catagen phase, which lasts approximately a few weeks, is the transition stage after the anagen phase. During the catagen phase, hair growth slows and the bulb of the hair begins to degenerate.

    59

  3. The telogen phase, when the hair has stopped growing, is the resting stage after the catagen phase. Hair in the telogen phase will remain in place for approximately one year, until the follicle initiates another anagen phase; at that point, the hair falls out. It is estimated that, at any given time, approximately 10 to 15 percent of head hair is in the telogen phase.

86. The duration of the three stages of growth varies across individuals and also within an individual. Different strands of hair grow at different rates, so no single strand will necessarily reflect the overall growth of hair during the period in question. When a pencil-sized collection of hair is gathered (as recommended), drugs present in the sample almost never form a discrete band. As the individual’s hair grows, the width of the band containing the drug-contained hair increases, as the hairs in the telogen phase remain static, while those in the anagen phase continue to grow. The widening of the band is a process referred to as “banding.” As a result of banding, it can take several weeks or even a few months before the hair closest to the scalp will test negative after a person has abstained from drug or alcohol use.

4.3.3 Sample Collection

87. Although sample collectors are instructed to cut the hair as close to the scalp as possible, closeness may depend on the experience of the collector as well as the type of hair involved (curly hair, for example, can be more difficult to cut closer to the scalp). In one study, researchers evaluated the amount of hair left on the donor’s scalp following collection by those having varying levels of experience. Even with samples collected by the more experienced collectors, the average length of hair remaining on the scalp after collection was 0.7 to 0.9 cm.30 Assuming an average rate of growth of 1 cm per month, about three to four weeks’ worth of hair left on the individual’s scalp (without the laboratory’s knowledge that three to four weeks’ worth was left) could affect the interpretation of the result by that amount of time.

4.3.4 Potential Residual Concentrations

88. Studies have suggested that, following chronic or heavy use of a drug, certain drugs such as cocaine may persist in tissues for at least a few months after the last use. As a result, it cannot be discounted that, following chronic or heavy use of a drug, tissue depots may result in low positive hair results for several weeks after discontinuation of drug use.

4.4 Potential for External Contamination for Drugs of Abuse

89. As noted, washing is recommended before testing hair samples and required if the test is to assess drug use, rather than exposure. However, studies have shown that, given the manner in which drugs incorporate into hair, even with the most aggressive washing procedures there is no way to rule out external contamination entirely.

90. The role of external contamination is a significant issue in the hair-testing field. In 2009, for example, the US Federal Bureau of Investigation’s (FBI’s) forensic laboratory suspended all cocaine analysis on hair, with the exception of criminal investigations into the exposure of children, because of concerns about the risk of external contamination leading to false positive results.31 In 2014, researchers at the FBI’s forensic toxicology laboratory proposed a multi-criteria wash protocol as a possible way to distinguish between use and external contamination.32 Plans are under way to implement this new wash protocol for cocaine testing in adult samples submitted to the FBI laboratory for analysis.

91. Currently, the SoHT recommends considering the possible role of external contamination for all test results, regardless of the washing procedure used. The SoHT has, however, recognized several tools that can be used to assist in distinguishing between active drug use (ingestion) and external contamination.

4.4.1 Interpreting Drug Metabolites

92. The identification of drug metabolites may assist in distinguishing between ingestion and external contamination. As early as 1997, the SoHT recommended that laboratories test for benzoylecgonine and cocaethylene, which are metabolites of cocaine, and for 6-AM, which is a metabolite of heroin. However, studies have shown that benzoylecgonine and 6-AM can be formed from cocaine and heroin, respectively, by spontaneous hydrolysis (a chemical reaction that may result when a compound is exposed to water). This hydrolysis can occur in different ways, including during the sample extraction process. Accordingly, the fact that benzoylecgonine is present in a sample does not automatically mean the individual consumed and metabolized cocaine. Nor does the presence of 6-AM necessarily confirm ingestion of heroin. As a result, in its 1997 consensus statement, the SoHT recommended that laboratories use hydrolysis controls and that they correct for hydrolysis in assessing the metabolite concentrations.

93. In its 1997 consensus statement, the SoHT recommended the use of a metabolite-to-drug ratio of 5 percent for benzoylecgonine-to-cocaine to interpret hair test results. Under this principle, if the concentration of benzoylecgonine is 5 percent or greater than the concentration of cocaine, the result will indicate active use rather than passive exposure. However, later published studies refuted the reliability of simply depending on the ratio to distinguish between use and external contamination. The SoHT’s later consensus statements for the examination of drugs of abuse in hair (dated 2004 and 2012) do not include metabolite-to-drug ratios for the interpretation of hair test results for drugs of abuse.

94. Ultimately, although metabolite-to-drug ratios may still assist in distinguishing between drug use and external contamination, as with the other aspects of hair testing there is no hard and fast rule. In addition, although benzoylecgonine is the most common metabolite of cocaine, other metabolites, such as norcocaine and cocaethylene, are now considered to be more reliable indicators of cocaine use because, unlike with benzoylecgonine, they are not susceptible to spontaneous creation as a result of hydrolysis.33

4.4.2 Analyzing Washings

95. Recognizing that an analysis of the solutions used to wash the hair (washings) can assist in assessing the potential role of external contamination, in its 2004 consensus statement the SoHT recommended that the washings be stored for later analysis, if necessary. The general rule is that the detection of 10 percent or more of the drug in the washings relative to the amount in the hair suggests there was external contamination that needs to be considered in the analysis. Ultimately, however, as with the other aspects of hair analysis, the interpretation depends on the circumstances, including the washing procedure used.

4.5 Dose–Response Relationship and Inter-subject Variation

96. Controlled dosage studies have supported an approximately linear relationship between dose and drug concentrations within an individual – suggesting that, when a person uses more of a drug, the concentration of the drug found in their hair increases by a corresponding amount. As a result, at least in theory, it may be possible to assess any increases or decreases in drug use or exposure by examining any increases or decreases in the concentration of drugs found in a person’s hair over time.34

97. However, a similar relationship between dose and response does not exist when considering hair test results from different individuals. The most likely explanation for the lack of a linear dose–response relationship is what has been called “inter-subject variation.” This term refers to the factors described above (including hair colour and use of cosmetic treatments), as well as to differences in how individuals metabolize drugs. Indeed, individual metabolism can have a significant impact on the amount of drugs or metabolites found in a hair sample – in studies where individuals with the same hair colour have consumed the same amount of drugs, the drug concentrations in their hair have still varied considerably. As a result, if two people are given the same dose of a drug, it is expected that the hair tests would result in different drug concentrations. Inter-subject variation is not unique to hair testing; blood and urine concentrations are also different between people given the same dose of a drug.

4.6 Segmental Analysis and Serial Testing for Drugs of Abuse

98. Although external contamination must always be considered in both analyses, the benefit of segmental analysis and serial testing (which refers to the practice of testing the same individual’s hair over time) is that they control for many of the variables that would otherwise affect the analysis. For example, differences across natural hair colour and other concerns with inter-subject variation are less relevant if the results are being compared for the same individual (assuming the individual has not undergone any significant hair treatments during the period being tested). As a result, segmental analysis and serial testing are both seen to be reliable methods to determine trends in drug use or exposure over time.

4.7 Infant- and Child-Specific Considerations for Drugs of Abuse

99. Several factors should be considered when interpreting infant or child hair tests that are positive for drugs or metabolites. First, infants and children are particularly sensitive to environmental exposure. Frequent handling by a caregiver who is a regular cocaine user may, for example, result in the transfer of drugs and metabolites to the child through the caregiver’s sweat. In addition, young children who are crawling or starting to walk may come into contact with drugs or drug residue from contaminated surfaces in the home. As a result, when interpreting an infant or child test for the purposes of assessing if the infant or child ingested the drug, it is particularly important to consider evidence of the primary caregiver’s drug use, as well as the environment at home.

100. Second, drugs or metabolites can also incorporate into hair in utero. Many newborns are born with hair of varying lengths and, although a great number lose their hair in the months after birth, some do not. If a child’s hair has not been cut since birth and the mother was known to use drugs during pregnancy, any drugs identified in the area of the hair farthest from the scalp may be attributed to exposure to drugs in the womb. Because growth rates vary considerably for infants and children, full details relating to the infant’s or child’s hair and age should be requested and the test results interpreted with caution. In addition, a newborn’s meconium may be tested for drugs of abuse.35 An infant’s or child’s hair test results should be interpreted in conjunction with meconium test results, if available. A positive meconium test would suggest gestational exposure to drugs; a negative one would not rule out gestational exposure but would suggest that the mother had not been regularly using drugs in the second and third trimesters.

101. Third, drugs can also be transferred through breast milk. If an infant is being breastfed by a mother who is known to be or suspected of using drugs, the interpretation of a positive hair test for the infant should take into account the possibility that the drug or metabolite that incorporated into the infant’s hair was from ingesting the mother’s breast milk, rather than through ingestion of the drug.

4.8 Interpreting Body Hair

102. What I have described until now relates to the testing of head hair. In some cases (e.g., when the individual is bald), a hair test may be carried out on other body hair, including underarm hair, pubic hair, and, for men, chest and beard hair. The testing of body hair carries additional complications, including different growth rates for the hair and different proportions of hair actively growing or in the resting phase. Where head hair is not available, pubic hair is sometimes considered a good alternative; however, consideration should be given to the greater potential for external contamination from urine. Given the factors adding to an already complicated analysis, non–head-hair results must be interpreted with great caution.

4.9 Interpreting Alcohol Markers

103. Unlike for hair analysis for drugs of abuse (which cannot, in and of itself, distinguish between use and exposure), the SoHT has accepted that the analysis of alcohol markers in hair can provide a historical assessment of alcohol use. Since 2009, the SoHT has provided cut-offs for testing for “chronic excessive alcohol consumption,” which according to the World Health Organization describes drinking in excess of 60 g of pure ethanol (or approximately four to five drinks) per day for several months. In 2014, the SoHT updated its consensus statement to also provide cut-offs to assess abstinence from alcohol. That said, there is still inter-subject variation that must be taken into account notwithstanding the use of the cut-offs recommended by the SoHT, and the cut-offs should thus be treated as guidelines only.

104. The science of alcohol marker testing has evolved over the past decade, as more research has been carried out on the reliability of FAEEs and EtG to assess the extent of a person’s alcohol use. This evolution has manifested itself in three ways:

  1. The SoHT’s cut-offs have evolved over time, depending on the length of the hair sample being tested.
  2. Given the effect of ethanol-containing hair products on FAEEs concentrations in hair, since the mid- to late 2000s it has become a widely accepted practice to obtain a list of the hair products used by the individual (and the ingredients of those products) to assess the potential impact of any ethanol-containing hair products on FAEEs test results.
  3. In view of the impact of ethanol-containing hair products on FAEEs analysis, since approximately the early 2010s, EtG analysis has become widely accepted as the more reliable alcohol marker to test for both abstinence and chronic excessive alcohol consumption. Many hair-testing experts do not consider FAEEs analysis to add anything to the EtG analysis, and many forensic laboratories have stopped testing for FAEEs altogether.

Given the recent developments in alcohol marker testing, what might have been an acceptable practice 10 years ago may no longer be one today.

4.10 Understanding the Limitations of Hair Analysis

105. Although hair analysis has the potential to provide a historical profile of drug or alcohol use or exposure, it is important to understand what hair analysis cannot do.

106. First, although segmental analysis can provide approximate dates of drug use or exposure, hair analysis cannot determine specific dates of use or exposure.

107. Second, hair analysis cannot determine the frequency of use or exposure. For example, it is not possible to differentiate between intermittent abstinence or weekend drug use and regular use. As a result, a hair test result cannot establish if the person tested is a regular or occasional drug user.

108. Third, a hair test cannot determine the precise dose of the drug used or to which the person was exposed.

109. Fourth, a hair test cannot determine if a drug was misused by a person (e.g., in the case of prescription drugs) or the level or extent of resulting impairment. Even for alcohol markers, for which the SoHT has established cut-offs to assess chronic excessive alcohol consumption, a hair test cannot determine if the person was or would have been impaired (or the extent of impairment) as a result of that level of consumption.

110. Fifth, in the child custody or child protection context, hair test results cannot assess whether a person consumed the drugs or alcohol when in a parental role or otherwise determine whether the person is fit to parent.

111. Ultimately, a hair test may be one tool that can be used to assess a person’s use or exposure to drugs or alcohol. It is only one piece of information and should be considered with all the other evidence. However, given the many factors that may affect a result, it is essential that hair tests be carried out and interpreted properly by experts who have the training and qualifications to do so.


Notes

15 See Appendix 6 for further reading on the science of hair testing.

16 Drugs and alcohol markers have different chemical structures and can be classified according to the functional groups they contain, which have acidic, basic, or neutral properties. Knowing whether the drugs or alcohol markers have acidic, basic, or neutral properties facilitates the optimization of extraction techniques for those drugs and alcohol markers as well as their analysis in biological matrices, including hair.

17 In addition, the eyelids, underarm, ear canal, and pubic region contain apocrine glands. These glands secrete sweat under the skin, from which drugs can also diffuse into the hair.

18 Cocaethylene is a metabolite that forms from cocaine when the individual also co-consumes alcohol.

19 It was not until 2009 that the SoHT released its first consensus statement containing its recommended guidelines on the examination of alcohol markers in hair.

20 In some cases, where there are concerns about leaving a visible bald patch (e.g., with small children or with adults who have thinning hair), smaller volumes may be collected from multiple sites within the same region.

21 In the hair strand depicted in Figure 3.2, for example, the drug, which was in the 1–2 cm section of the hair, might not be detected if the 12 cm strand was tested in bulk because the greater volume of hair would dilute the amount of the drug. On the other hand, if the 12 cm strand was tested in 3 cm segments, there is a greater likelihood of detecting the drug in the 0–3 cm segment since there would be a greater amount of the drug relative to the amount of the hair being tested.

22 The distinction between screening tests and confirmation tests, as well as use of the term “preliminary positive,” is described more fully below.

23 In fact, chopping is the recommended method for FAEEs analysis (although powdering is the recommended technique for EtG testing).

24 This colour change also makes it easier for the instrument to read the optical density.

25 As a result of the way in which the ELISA test works, only one drug or metabolite can be tested on the sample at any given time.

26 See T. Mieczkowski and R. Newel, “Statistical Examination of Hair Color as a Potential Biasing Factor in Hair Analysis” (2000) 107(1–3) Forensic Science International 13–38.

27 D.A. Kidwell, F.P. Smith, and A.R. Shepherd, “Ethnic Hair Care Products May Increase False Positives in Hair Drug Testing” (2015) 257 Forensic Science International 160–4.

28 See G. Cooper, “Anatomy and Physiology of Hair, and Principles for Its Collection,” in P. Kintz, A. Salomone, and M. Vincenti, eds, Hair Analysis in Clinical and Forensic Toxicology (Amsterdam: Elsevier / Academic Press, 2015) 1–22, for a summary of published growth rates for different types of human hair.

29 For example, one study has shown that Caucasian hair tends to grow faster than Asian hair. As well, hair can grow faster in women than in men. In addition, the rate of hair growth (and growth generally) tends to decline with age: M.R. Harkey, “Anatomy and Physiology of Hair” (1993) 63(1–3) Forensic Science International 9–18.

30 M.A. LeBeau, M.A. Montgomery, and J.D. Brewer, “The Role of Variations in Growth Rate and Sample Collection on Interpreting Results of Segmental Analyses of Hair” (2011) 210(1–3) Forensic Science International 110–16.

31 See M.A. LeBeau and M.A. Montgomery, “Considerations on the Utility of Hair Analysis for Cocaine” (2009) 33(6) Journal of Analytical Toxicology 343–4.

32 C.L. Morris-Kukoski, M.A. Montgomery, and R.I. Hammer, “Analysis of Extensively Washed Hair from Cocaine Users and Drug Chemists to Establish New Reporting Criteria” (2014) 38 Journal of Analytical Toxicology 628–36.

33 However, norcocaine has been detected in low quantities in some cocaine batches, suggesting that it is not created only from the metabolism of cocaine in the body; and cocaethylene has been shown to be present in street cocaine as a result of formation during the purification stage by illicit manufacturers.

34 At the same time, however, the interpretation of hair test results requires consideration of a variety of factors, and a linear relationship between dose and response within an individual cannot be assumed.

35 Meconium is the fecal material that accumulates in the fetus’s intestines and is discharged at or near the time of birth.

Chapter 4
Structure and Evolution of MDTL

1. Introduction

1. Before I discuss the adequacy and reliability of the testing methodology used by the Motherisk Drug Testing Laboratory (MDTL or the Laboratory), it is important, first, to situate MDTL within the Hospital for Sick Children (SickKids or the Hospital) and, second, to describe the structure of the Laboratory and the way it developed over the past three decades. During that time, MDTL evolved from carrying out research on neonatal hair analysis to offering a forensic hair-testing service for use primarily in child protection, but also in criminal, cases.36 By 2005, the start of the period under review, the Laboratory was performing hair tests on well over a thousand samples annually for those purposes.37

2. Situating MDTL within the Hospital for Sick Children

2. SickKids has a complex structure. It is organized into departments, divisions, programs, and services, as well as the separate Research Institute, none of which forms a vertical hierarchy. Of particular relevance to MDTL are the Motherisk Program, the Research Institute, the Department of Paediatric Laboratory Medicine, and the Division of Clinical Pharmacology and Toxicology.38

2.1 The Motherisk Program

3. Founded in 1985, the Motherisk Program at the Hospital provides information and guidance about the potential risks to a developing fetus or infant from exposure to drugs, chemicals, diseases, radiation, and environmental agents. The program describes itself as a clinical, research, and teaching program dedicated to antenatal drug, chemical, and disease risk counselling. It currently offers a variety of services to members of the public and physicians, including telephone and in-person counselling to pregnant and breastfeeding women and caregivers about the safety of maternal medications, naturopathic products, and other exposures, as well as a helpline about alcohol and substance abuse.

2.2 The Research Institute

4. The Research Institute was founded in 1954 and is currently one of the largest hospital-based research institutes in Canada. Research activities are coordinated under seven programs: cell biology, child health evaluative sciences, developmental and stem cell biology, genetics and genome biology, molecular structure and function, neuroscience and mental health, and physiology and experimental medicine. The research that is carried out at the Hospital is done through a “principal investigator” – the person with primary responsibility for the individual research project. Dr. Gideon Koren’s research fell largely within the Research Institute’s child health evaluative sciences program.39 MDTL was considered to be a research laboratory within the Research Institute until the fall of 2014, when formal oversight of its operations was transferred to the Department of Paediatric Laboratory Medicine.

2.3 The Department of Paediatric Laboratory Medicine

5. The Department of Paediatric Laboratory Medicine (DPLM) was established in 1997 and operates under regulatory and accreditation compliances required by the Laboratory and Specimen Collection Centre Licensing Act (Laboratory Licensing Act).40 It houses SickKids’s clinical laboratories and comprises five divisions: pathology, biochemistry,

hematological pathology, microbiology, and molecular genetics. As a toxicology laboratory, MDTL fell within DPLM’s biochemistry division.

2.4 The Division of Clinical Pharmacology and Toxicology

6. The Division of Clinical Pharmacology and Toxicology (the Division) was established in 1979 within the Department of Paediatrics. It focuses on drug-related issues in pediatrics, including adverse drug reactions, drugs of abuse, overdoses, and poisoning. The Division supports the Hospital’s clinical pharmacology and toxicology activities, including the Motherisk Program. Other clinical activities that the Division supports include a consultation service and the Ontario Poison Centre. For postgraduate education, the Division is one of three major sites for the Clinical Pharmacology Residency program offered by the University of Toronto. Graduate students and fellows from the Division carried out research at MDTL.

3. Structure of MDTL

7. The leadership of MDTL throughout the period under review comprised three positions: laboratory director, laboratory manager, and, starting in October 2009, laboratory supervisor. In addition, MDTL’s staff included laboratory technicians and, after October 2009, laboratory technologists; a research associate and research technologists; and beginning in June 2008, a quality manager. In late 2008, MDTL also began to use graduate students and fellows in the Division, who were not permanent laboratory staff, to provide over-the-phone interpretations to customers of the hair test results the Laboratory generated. An overview of each of these positions is provided below.

3.1 Laboratory Director

8. Dr. Koren is a physician and clinical toxicologist with a long career at SickKids. He joined the Hospital in 1982 and was the founder and director of the Motherisk Program throughout the period under review. He has sat on the editorial boards of numerous academic journals, held several academic appointments (including with the University of Toronto), and been involved in hundreds of research studies. However, Dr. Koren has never had any formal training in forensic toxicology or any experience in a forensic toxicology laboratory.

9. Dr. Koren was officially appointed director of MDTL in early 2009. However, even before then, he referred to himself as the director of MDTL and was the person with ultimate responsibility for the Laboratory.41 Although he was responsible for the research team at the Laboratory, Dr. Koren did not actually perform the hair tests, and he rarely provided interpretations of MDTL’s results in individual cases. He told the Independent Review that he was not the technical expert with respect to the development and performance of the Laboratory’s tests; rather, he recruited competent technicians and technologists to develop the testing, delegated to people he trusted, and expected to be asked if anyone had a question or concern. Even so, Dr. Koren signed the Laboratory’s interpretation reports for a brief period following the departure of MDTL’s first laboratory manager in April 2005, and he was called to testify in court from time to time (including at the Broomfield trial in 2009).42

3.2 Laboratory Manager

10. MDTL’s laboratory manager from June 1995 to April 2005 was Julia Klein.43 Ms. Klein holds a master of science degree in organic chemistry. She joined SickKids in 1979 as a laboratory technician and began working with Dr. Koren shortly after he joined the Hospital in 1982.

11. Ms. Klein did not have any training in hair analysis before she arrived at SickKids. However, after she joined the Hospital, she had extensive communications with the founder of Psychemedics, a laboratory in California, regarding drug extraction procedures from hair, radioimmunoassay, and the interpretation of hair test results. She also participated annually, starting in 1996, in almost all the workshops and meetings organized by the Society of Hair Testing (SoHT), which was (and is) the only international organization dedicated to the development and standardization of hair-testing methodologies. Moreover, in or around 2000 or 2001, Ms. Klein attended a training session hosted by Immunalysis Corporation,44 and in 2002 she went to Professor Fritz Pragst’s laboratory at the Institute of Legal Medicine and Forensic Sciences, University Hospital Charité, in Berlin to learn the methodology for hair testing for fatty acid ethyl esters (FAEEs).45

12. Ms. Klein was instrumental in developing the testing methodology that MDTL used from 2005 to August 2010 to test for drugs of abuse, including the use of ELISA to quantify drug concentrations in hair. Moreover, the concentration ranges that formed the basis of MDTL’s interpretation opinions were first created and implemented during Ms. Klein’s tenure as laboratory manager.

13. Ms. Klein left MDTL suddenly on April 29, 2005. The Hospital told the Independent Review that, in April 2005, it discovered that a senior secretary working at the Research Institute (who was responsible for the receipt of funds for MDTL’s hair tests) had allegedly misappropriated funds from MDTL’s customers. Dr. Koren told the Independent Review that, because of this discovery, he was asked to let Ms. Klein know that a forensic accounting for all of MDTL would be initiated. As a result of these investigations, SickKids terminated the employment of both the senior secretary and Ms. Klein purportedly for cause.46 Subsequent to their departures, MDTL’s accounting system was changed.

14. After Ms. Klein’s departure, Dr. Koren appointed Joey Gareri as acting laboratory manager to replace her, and in September 2006 his appointment was made official.47 At the time, Mr. Gareri was a research trainee at MDTL and was a master of science student in clinical pharmacology and biomedical toxicology.48 His research involved evaluating the incidence of fetal alcohol exposure by analyzing meconium samples for alcohol metabolites. Having worked at SickKids through his student days, Mr. Gareri had been supervised by Dr. Koren and looked to him as his mentor.

15. At the time of his appointment, Mr. Gareri had no formal training or experience in forensic toxicology or hair analysis and no management training. As a result, he sought out and gained knowledge about hair testing, for example, by joining and attending SoHT conferences, attending weekly rounds at MDTL and toxicology rounds at the Ontario Poison Centre, and going to conferences hosted by various organizations, including the International Association for Therapeutic Drug Monitoring and Clinical Toxicology and the International Association of Forensic Toxicologists.

16. Like Dr. Koren, Mr. Gareri did not perform the hair tests. In fact, Mr. Gareri told the Independent Review that, although he understood the principles underlying immunoassay-based tests such as ELISA, he had never performed an ELISA test. Starting in November 2005, Mr. Gareri signed all the Laboratory’s interpretation reports and had primary responsibility for providing interpretations of MDTL’s test results to customers – the majority of whom were child protection workers. Beginning sometime in 2006, Mr. Gareri also began to testify in court about MDTL’s hair test results, including at the Broomfield preliminary inquiry in 2007.49

17. Ultimately, in addition to the financial, administrative, and managerial tasks Mr. Gareri was given, his primary role was to communicate the test results to the Laboratory’s customers and other users – both inside and outside the courtroom.50

3.3 Laboratory Supervisor

18. MDTL did not originally have a laboratory supervisor. However, in 2008 it was asked to participate in the Hospital’s next accreditation inspection under the Ontario Laboratory Accreditation program, which was then scheduled to take place in 2010 (but was ultimately held in January 2011). In October 2009, in the lead-up to MDTL’s accreditation inspection, MDTL appointed Dr. Bhushan Kapur as laboratory supervisor.51

19. Dr. Kapur is a pharmacologist, toxicologist, and biochemist. He worked at the Addiction Research Foundation laboratories for 24 years, from 1971 to 1995, before joining SickKids as a scientist / consultant toxicologist in the Division. Although Dr. Kapur had also worked as a consultant for a number of other clinical laboratories, he had no experience with hair testing. In the lead-up to MDTL’s first accreditation inspection, Dr. Koren asked Dr. Kapur to come to MDTL to assist in the Laboratory’s accreditation process. Dr. Kapur agreed, and starting in October 2009, he became MDTL’s first laboratory supervisor. His role as laboratory supervisor was part-time, as he continued on as a scientist / consultant toxicologist in the Division.

20. Dr. Kapur’s focus when he joined MDTL in October 2009 was on method development for the Laboratory’s new gas chromatography–mass spectrometry (GC-MS) instrument and on overseeing the development of MDTL’s standard operating procedures. After the Laboratory had transitioned to GC-MS for some hair tests and developed its standard operating procedures, in December 2010, Dr. Kapur’s role at MDTL consisted primarily of reviewing quality control data generated by the Laboratory, answering questions, and reviewing any issues that were identified in its quality management system. Other than his review of quality control data and general trouble-shooting, Dr. Kapur was not involved in the Laboratory’s analytical testing, its interpretations, or its communications with customers.

3.4 Other Staff

21. Apart from laboratory management, MDTL employed other staff who were involved in the analytical testing carried out at the Laboratory.52

3.4.1 Laboratory Technicians and Laboratory Technologists

22. The laboratory staff who carried out MDTL’s hair tests for alcohol markers and drugs of abuse were laboratory technicians and laboratory technologists.53 Until August 2010, all the steps in the analytical process for drugs of abuse – from sample preparation through to the testing procedure – were carried out by MDTL’s laboratory technicians. Although others assisted with sample preparation, one laboratory technician in particular, Tatyana Karaskov, was responsible for carrying out virtually all MDTL’s ELISA tests during the period under review.

23. Ms. Karaskov was trained as a physician and had 11 years of experience working in laboratories before going to MDTL. She joined SickKids in 1998, first as a volunteer and, six months later, as an employee. She told the Independent Review that, when she first started at MDTL, she was working with fluids, not hair. She could not recall when she began carrying out the hair tests for drugs of abuse, but said she was trained by Ms. Klein and another laboratory technician on how to perform the tests.

24. In the lead-up to MDTL’s accreditation inspection in early 2011, MDTL hired two laboratory technologists. In October 2009, the first laboratory technologist began to carry out MDTL’s FAEEs analysis (previously, this analysis had been done by a laboratory technician or a research technologist). In September 2010, after MDTL changed its testing methods to routine GC-MS confirmation of its tests for drugs of abuse, a second laboratory technologist, Paula Walasek, became responsible for carrying out this analysis.54

25. Ms. Walasek is a member of the College of Medical Laboratory Technologists of Ontario.55 She had previously worked in a laboratory, carrying out GC-MS analysis (though not on hair), before joining SickKids’s Brain Tumour Research Centre in 2006. In October 2008, MDTL first hired Ms. Walasek, as a research technologist, to assist with the development of MDTL’s GC-MS procedure. Once the new procedure was implemented in September 2010, her role became that of a laboratory technologist, and she was given primary responsibility for carrying out the hair tests under the GC-MS procedure that she had assisted in developing. After September 2010, Ms. Karaskov continued to perform certain hair tests using ELISA.

26. Although MDTL’s laboratory technicians and technologists carried out the hair tests, they were not responsible for communicating the results to customers. As described above, that task fell to the laboratory manager and, as set out below, starting in late 2008, also to graduate students and fellows at the Laboratory.

3.4.2 Research Associates and Research Technologists

27. After a new instrument has been acquired and commissioned, the technical aspects of the method must be validated before it may be used for clinical or forensic purposes. Validation requires continual testing on the instrument to determine such factors as the minimum sample volume needed (to determine the instrument’s sensitivity), the number of calibrators required to generate a reliable calibration curve, the limits of detection and quantitation of the instrument, the cut-off to be applied for the particular drug or metabolite, and the manner in which the instrument identifies and quantifies the drug or metabolite.

28. MDTL employed a research associate and research technologists to assist with instrumentation validation and method development at the Laboratory. The first and only research associate was Dr. Katarina Aleksa, who has a doctorate in pharmacology. After joining MDTL in November 2006, she worked on implementing a method for the analysis of hair for FAEEs at the Laboratory. In October 2007, MDTL hired Netta Fulga (who would later become MDTL’s quality manager) as a research technologist to assist with the development of the FAEEs method and to carry out FAEEs analysis once the method was fully implemented. While a research technologist, Ms. Fulga also developed MDTL’s first standard operating procedures for FAEEs analysis. These procedures became the template for the standard operating procedures that MDTL later developed for drugs of abuse.

29. In the first few years of Mr. Gareri’s tenure as laboratory manager, MDTL relied on graduate students to work up a GC-MS method for drugs of abuse as part of their graduate research. However, in October 2008, in the lead-up to the accreditation inspection, and recognizing that reliance on graduate students to develop a GC-MS method would take too long, MDTL hired Ms. Walasek as a research technologist to assist with method development for drugs of abuse. Once the GC-MS method was implemented, Ms. Walasek carried out the tests. Both Dr. Aleksa and Ms. Walasek worked on the development of the standard operating procedures for MDTL’s hair tests for drugs of abuse, which came into effect in late 2010.56

3.4.3 Quality Manager

30. In June 2008, MDTL hired a quality manager, Netta Fulga, to create, implement, and oversee the Laboratory’s quality management system. Ms. Fulga has a master of science degree in biochemistry. When she accepted the position of quality manager, she had already been employed as a research technologist at MDTL for over eight months. In December 2011, after appropriate training, Ms. Fulga became a certified quality auditor with the American Society for Quality.57

31. Under the oversight of Dr. Kapur, Ms. Fulga implemented MDTL’s quality management system, including a set of standard operating procedures and quality control protocols for the hair tests. Following its implementation, Ms. Fulga’s role was to oversee and maintain the quality management system, including by review of the quality control data for the tests that MDTL’s laboratory technicians and laboratory technologists carried out. Ms. Fulga played no role in the interpretation of MDTL’s hair test results for customers.

3.5 Graduate Students and Fellows

32. Graduates and fellows from the Division carried out research at MDTL. The graduate students and fellows rotated through the Laboratory as they completed their studies or their fellowships.

33. Starting in late 2008, and because of the increasing number of tests at MDTL, the Laboratory began to use graduate students or fellows (whom MDTL called “laboratory counsellors”) to assist in providing interpretations of the Laboratory’s hair test results.58 They had no formal training in forensic toxicology or hair analysis and were not permanent members of MDTL staff. Nevertheless, they were given responsibility for providing what MDTL called “routine” interpretations over the telephone to customers who called to ask about the test results.

34. This model of providing over-the-phone interpretations appears to have developed from the telephone counselling that the Motherisk Program offered to pregnant and breastfeeding women.

4. Understanding Forensic vs. Clinical vs. Research Purposes

35. MDTL started out as a research laboratory within the Hospital’s Research Institute. Although it eventually participated in the clinical laboratory accreditation process with the other clinical laboratories in the Hospital, throughout the period of the Review MDTL provided hair-testing services for forensic purposes. It is therefore important to understand the distinctions among the uses of hair tests for research, clinical, and forensic purposes.

  • Hair tests for research purposes are carried out to advance the science of hair testing – for example, to develop or improve hair-testing methodologies or to assist in the interpretation of hair test results (e.g., research on the impact of washing procedures on the ability to remove external contaminants from hair samples or the effect of natural hair colour on the incorporation of drugs into hair). They do not usually have legal consequences, nor do they have any clinical purpose.
  • Clinical tests are intended to be used for diagnostic, prophylaxis, or treatment purposes. Examples of clinical purposes in the hair-testing context include determining if a person had particular drugs in her or his system in order to establish an immediate treatment plan. In the longer-term context, the treatment plan might include monitoring relative changes in a person’s drug intake for compliance purposes (e.g., to assess if a person is still using prescribed drugs when treatment does not appear to be working).
  • Forensic science involves the application of scientific knowledge and methodology to legal problems. However, a forensic test is not necessarily one that will be used in the justice system; a test will be forensic if it is carried out or used for a legal (or medico-legal) purpose. In the hair-testing context, medico-legal purposes include detecting use of or exposure to drugs for the purposes of child protection or family law cases (e.g., as evidence of drug use by a parent or child), as well as in criminal proceedings (e.g., as evidence of whether a complainant, accused, deceased, or other witness was drugged). Other forensic (or medico-legal) purposes that may not necessarily involve a courtroom at first instance, but may nonetheless have legal implications, include workplace drug testing or sports testing for banned substances.

36. One may not be able to draw a bright line between these purposes in all circumstances – tests that were originally run for clinical purposes may later have legal implications. For example, a child protection agency may become involved if a hair test on a newborn that was originally carried out to determine gestational exposure to cocaine for treatment (i.e., clinical) purposes turns out to be positive. In such circumstances, there can be a blurring of the lines between the clinical and the forensic worlds. In some cases, it may be appropriate to use a clinical test for forensic purposes with that caveat; in others, it may be more appropriate to rerun the test in a forensic setting before relying on it for forensic purposes.

37. It is important to distinguish among research, clinical, and forensic purposes because different standards apply, depending on the manner in which the hair test is intended to be used. Forensic laboratories have the most stringent requirements, including strict chain-of-custody procedures and documentation standards to reduce the risk of contamination of samples and to ensure the traceability and reliability of the results obtained.

38. Although clinical laboratories employ many standards and procedures that are similar to those used in forensic laboratories, chain of custody is not required or expected in clinical cases. Indeed, chain of custody, as a concept, would be unworkable in a clinical setting where a quick turnaround is often essential, given the potential impact of the laboratory test on diagnostic and treatment decisions.

39. As for research purposes, it is usually up to the researcher (or principal investigator) to develop the procedures to be followed, to validate those procedures, and ultimately to report the methods used, together with the study’s findings, in any published articles. That said, it is still important that the results obtained in any research study are accurate and reliable, as observations from research can be, and often are, applied to both medico-legal and clinical cases.

5. Evolution of MDTL’s Hair-Testing Work

40. Shortly after the Motherisk Program was founded in 1985, researchers at MDTL began to publish studies on neonatal hair analysis for drugs of abuse. In particular, in 1989 researchers published their first peer-reviewed article, in the Journal of the American Medical Association, describing the ability to detect gestational exposure to cocaine through neonatal hair analysis.59 It was followed, in 1993, by an article in the New England Journal of Medicine describing the ability of neonatal hair analysis to detect gestational tobacco exposure.60 Over the next decade, researchers at MDTL continued to publish their research on the ability to detect drugs or other substances through hair analysis.

41. SickKids is one of Canada’s most research-intensive hospitals. In the late 1980s and early 1990s, the science of hair analysis was in its infancy. MDTL was seen to be on the cutting edge of the science, and by the late 1990s it began to hold itself out as a leader in the field of hair testing for drugs of abuse.61

42. As MDTL became more prominent in the field, there was a shift in the Laboratory’s work – it moved from carrying out hair tests purely for research purposes to offering hair-testing services to customers, primarily for use in child protection proceedings (but in criminal cases as well). Ms. Klein, MDTL’s laboratory manager at the time, told the Independent Review that MDTL started to conduct hair tests for use in child protection or criminal cases around 1999, although she could not recall the exact year. When asked what led MDTL to begin this service, she explained that the Laboratory was originally contacted by child protection agencies and lawyers with requests to test hair for cocaine. The Hospital also told the Independent Review that, in 1999, “likely as a result of the research publications, the MDTL began receiving regular requests from Children’s Aid Societies requesting that samples be tested and reported to assist in the discharge of their responsibilities under the child welfare and protection legislation.” Likewise, after MDTL published articles on hair testing for drugs, Dr. Koren said that the demand for hair tests from child protection agencies “developed organically.”

43. In 1999, Ms. Klein, Ms. Karaskov, and Dr. Koren submitted an article to Forensic Science International (published in 2000) on the applications of hair analysis for drugs of abuse.62 The authors reported that, from October 1991 to April 1999, they had analyzed more than 1,000 hair samples for drugs of abuse, with the number increasing each year during that period. They described two of the “clinical applications” of hair testing as follows:

3.2 Children’s exposure to drugs in the home

We are often called upon to test children’s hair for different drugs they may be exposed to in the home. This happens usually in the context of child custody cases when one parent may accuse the other of using drugs and exposing the child passively (through secondhand smoking) and/or actively (to sedate the child or as part of other abuse patterns).

3.5 Criminal cases

In three instances our expertise was called upon to verify cocaine use in people who committed one or multiple crimes. In all three cases, we detected large amounts of cocaine in the respective individual’s hair. In all cases [radioimmunoassay] was used and confirmatory analysis by GC–MS was performed.

One of these cases was tried in an Ontario court, and to the best of our knowledge it was the first time that hair drug test data were admitted in a Canadian criminal court.63

44. Accordingly, by 1999 MDTL had shifted from carrying out hair tests for drugs of abuse for research purposes to routinely carrying out those tests for customers for use in child protection and criminal cases. Several years later, in 2004, MDTL also began to offer hair-testing services to detect FAEEs in order to assess the extent of an individual’s alcohol consumption over time.

45. In the early 2000s, representatives of MDTL also began making presentations, primarily to child protection agencies, about the advantages of hair testing for drugs of abuse, including its ability to detect and quantify drug use over time. Ms. Klein told the Independent Review that, from approximately 2001 to 2004, she gave about 10 presentations to workers from different child protection agencies about hair testing generally. These presentations contained information comparing the usefulness of hair testing to urine testing because, Ms. Klein explained, child protection workers wanted help in choosing between the two methods for detecting drugs of abuse.

46. In May 2005, after Ms. Klein left MDTL, Mr. Gareri continued to give presentations about MDTL’s hair analysis. From May 2005 until April 2015 (when MDTL shut down its non-research operations), Mr. Gareri gave more than 150 presentations to various organizations across the province. Most of the organizations were child protection agencies, but he also presented to hospitals, to policing organizations, at legal conferences, and to the courts.64

47. The number of hair samples that MDTL tested for drugs of abuse and alcohol markers increased throughout the years 2005–15.65 In the years under review, MDTL tested an average of 2,000 samples a year at the request of Ontario child protection agencies.

48. Ultimately, the evolution in the nature of MDTL’s work provides important context as to how MDTL came to offer forensic hair-testing services without complying with forensic standards. By 2005, the start of the period under review, what had begun at MDTL in the late 1980s as research on the potential for neonatal hair analysis to detect gestational exposure to cocaine had morphed into a forensic service that the Laboratory offered to its customers – primarily for use in child protection but also in criminal proceedings.


Notes

36 MDTL also offered hair-testing services to members of the public, including parents involved in custody and access disputes in family law cases. The Independent Review’s mandate does not extend to such cases.

37 See Appendix 7 for Hair Testing Data, 2005–15, provided by the Hospital.

38 MDTL is one facet of the Motherisk Program.

39 Dr. Koren’s role at MDTL is described below.

40 Laboratory and Specimen Collection Licensing Act, RSO 1990, c L.1.

41 Dr. Koren accepted the position of director of MDTL on February 24, 2009. However, it is clear from documents obtained by the Independent Review that he referred to himself as director of MDTL even before his official appointment. For example, in an affidavit that Dr. Koren used in a child protection proceeding dated May 2005, he described himself as “Director of the Motherisk Laboratory.” For ease of reference, I refer to Dr. Koren as MDTL’s director throughout the period under review.

42 Dr. Koren also provided an expert report on the Broomfield appeal in January 2013.

43 Ms. Klein told the Independent Review that, even before 1995, she filled the role of de facto laboratory manager, including the supervision of laboratory work by the graduate students and laboratory technicians at MDTL.

44 Immunalysis Corporation is the manufacturer of the enzyme-linked immunosorbent assay (ELISA) kits that MDTL used during the period under review.

45 Professor Pragst is widely recognized as a leader in FAEEs analysis in hair.

46 Ms. Klein declined to answer the Independent Review’s questions about the circumstances surrounding her departure from the Hospital because, she said, she did not consider them to be relevant. However, Ms. Klein’s sudden departure from the Hospital explains why there was no transition from her to Mr. Gareri.

47 Although Mr. Gareri was acting manager for a period before he was officially appointed to the position, for ease of reference I refer to him as MDTL’s laboratory manager from May 2005 until MDTL shut down its non-research operations in April 2015.

48 Mr. Gareri received his master of science degree in 2006 and went on to obtain his doctorate in March 2015.

49 Mr. Gareri told the Independent Review that he was not comfortable testifying at the trial in 2009, and, as a result, Dr. Koren testified. Mr. Gareri provided two reports in 2013 for the Broomfield appeal.

50 For example, the job description that MDTL created for the laboratory manager position in the lead-up to the January 2011 accreditation inspection included specific responsibilities for issuing “written interpretations of laboratory results” and “providing objective expertise in writing or by personal appearance (i.e. court testimony) in support of MDTL laboratory results.”

51 “Laboratory supervisor” is a defined position both in a regulation to the Laboratory Licensing Act and in the Ontario Laboratory Accreditation requirements. Laboratory supervisors are required to hold certain qualifications. See Laboratories, RRO 1990, Reg 682, ss 1 and 6(2), and the Ontario Laboratory Accreditation requirements, 2013.

52 The roles described below focus on the hair analysis that MDTL carried out for use in child protection and criminal proceedings. I have not described the roles of those individuals who had purely administrative functions, including, for example, the Laboratory’s administrative assistants.

53 MDTL also used the term “medical laboratory technologist” to describe its laboratory technologists. Medical laboratory technologists (MLTs) are one of Ontario’s regulated health professions. They are regulated by the College of Medical Laboratory Technologists of Ontario. All of MDTL’s laboratory technologists were also MLTs, but for ease of reference I refer to them as laboratory technologists.

54 The Laboratory Licensing Act defines both laboratory technician and laboratory technologist. In addition, the Ontario Laboratory Accreditation requirements mandate that laboratory testing be completed by a laboratory technologist, or if completed by a laboratory technician, that it be supervised directly by one of a list of professionals, including laboratory technologists, supervisors, and directors.

55 In addition to obtaining her medical laboratory technology diploma, Ms. Walasek also has a bachelor of science degree and a master of science degree.

56 When Dr. Kapur joined MDTL in October 2009, he oversaw the work of Dr. Aleksa and Ms. Walasek in respect of both method development and the development of the Laboratory’s standard operating procedures. In September 2011, after MDTL acquired a liquid chromatography–tandem mass spectrometry (LC-MS/MS) instrument, Dr. Aleksa again worked on method development under the guidance of Dr. Kapur. The LC-MS/MS method was implemented for use in child protection and criminal cases in May 2014.

57 The American Society for Quality, a body with headquarters in the United States, provides training and professional certifications in quality management. Its scope covers a broad range of disciplines and is not limited to quality management in laboratories.

58 In addition, starting in 2011, the graduate students and fellows prepared written interpretation reports based on past reports and templates. They did not, however, sign the reports or provide any other form of written interpretation to customers.

59 K. Graham et al, “Determination of Gestational Cocaine Exposure by Hair Analysis” (1989) 26(23) Journal of the American Medical Association 3328–30.

60 J. Klein, D. Chitatyat, and G. Koren, “Hair Analysis as a Marker for Fetal Exposure to Maternal Smoking” (1993) 328(1) New England Journal of Medicine 66–7.

61 For example, in October 1998, MDTL hosted a two-day Second International Hair Testing Workshop, attended by toxicologists and physicians from around the world. In a fall 1998 newsletter reporting on the workshop, MDTL described itself as “Canada’s only academic laboratory to conduct routine analysis of neonatal and adult hair.” The newsletter reported that, at the workshop, “[p]anelists also discussed the application of hair analysis in Canadian criminal and civil law, as well as its utility in child protection and custody cases.”

62 J. Klein, T. Karaskov, and G. Koren, “Clinical Applications of Hair Testing for Drugs of Abuse – the Canadian Experience” (2000) 107(1–3) Forensic Science International 281–8.

63 In 1999, when the authors submitted this article for publication, MDTL was not routinely confirming its samples using GC-MS. The manner in which researchers represented MDTL’s testing methodologies in academic articles from 1999 through to 2010 is described in Chapter 8.

64 Mr. Gareri presented on a range of topics. He was unable to identify for the Independent Review the topics for all the presentations he gave during this period. Many presentations were on hair analysis for drugs and alcohol markers, and others were on addictions more generally as well as on fetal alcohol spectrum disorder.

65 MDTL’s financial forecast showed that its income for 2004/5 was $624,875. Based on the increasing number of tests, the Laboratory’s income would likely have increased also.

Chapter 5
Adequacy and Reliability of
Analytical Methods Used by MDTL

1. Introduction

1. In this chapter, I report on the adequacy and reliability of the analytical testing methodology that the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) used from 2005 to 2015. I also report on the extent to which MDTL met internationally recognized forensic standards during this period.

2. I begin with a discussion of the different sources that make up the “internationally recognized forensic standards” according to which MDTL’s practices are being assessed. For the reasons set out below, MDTL’s analytical methods did not meet these standards at any time from 2005 to 2015. Nevertheless, because MDTL’s methods for drug testing changed over time, this chapter is organized according to two time periods: 2005 to August 2010; and September 2010 to April 2015. MDTL’s methods for alcohol markers were different, so I cover the Laboratory’s tests for them separately.

3. In the 2005–August 2010 period, MDTL’s analytical practices for drugs of abuse suffered from two fatal flaws: first, the Laboratory relied on the unconfirmed results of its enzyme-linked immunosorbent assay (ELISA) tests – a preliminary screening test – both qualitatively (to distinguish positive from negative) and quantitatively (to calculate the drug concentration in the sample); and second, it had no written standard operating procedures for the hair tests it carried out, thereby calling into question the reliability and standardization of the analytical techniques it used. These two deficiencies alone are sufficient to render MDTL’s hair tests inadequate and unreliable during this period. However, the Independent Review identified additional ways in which the Laboratory’s analytical methods fell short of internationally recognized forensic standards, including

  • the absence of any oversight of the laboratory technician who was assigned to carry out the ELISA tests, with the result that there were significant reporting anomalies or errors in the case files reviewed by the Independent Review;
  • the failure to wash hair samples routinely before analysis;
  • inadequate chain-of-custody procedures; and
  • inadequate record keeping.

4. Starting in September 2010, MDTL began routinely to confirm the majority of its test results for drugs of abuse, initially using gas chromatography–mass spectrometry (GC-MS) and, after May 2014, liquid chromatography–tandem mass spectrometry (LC-MS/MS). MDTL also improved many of its laboratory practices by adopting written standard operating procedures and a quality management system. However, even in the September 2010–April 2015 period, MDTL continued to fall short of internationally recognized forensic standards by

  • relying on a flawed GC-MS procedure;
  • continuing to rely on unconfirmed ELISA results for certain drugs and for neonatal samples (though MDTL stopped reporting the quantitative results generated by those tests);
  • continuing to test unwashed hair samples from children and neonates; and
  • not maintaining appropriate chain of custody of the samples it tested.

5. Finally, although MDTL’s hair-testing methodology for fatty acid ethyl esters (FAEEs), at least from 2007 to April 2015, did not suffer from the same deficiencies as its testing methods for drugs of abuse, MDTL’s FAEEs tests were still inadequate and unreliable for forensic purposes: first, because of deficiencies in its GC-MS procedures and, second, because of its misapplication of cut-offs to assess alcohol consumption. Additionally, MDTL’s use of non-standard hair lengths, as well as its reliance on FAEEs analysis alone to assess alcohol consumption, though acceptable at the time, are now considered to be unreliable, given the evolution in the science. MDTL did not test hair samples for ethyl glucuronide (EtG). From February 2011 until the Laboratory closed, it sent all samples for EtG tests to the US Drug Testing Laboratory (USDTL) in Des Plaines, Illinois. The Independent Review has not assessed the USDTL hair-testing methodology, but MDTL also misapplied the cut-offs to assess alcohol consumption with respect to EtG.

2. Understanding Internationally Recognized Forensic Standards

6. My mandate requires me to review and report on “the extent to which the operation of [MDTL] between 2005 and 2015 was consistent with internationally recognized forensic standards.” Before I report on MDTL’s analytical methods, it is important to understand the specific standards to which MDTL’s practices are being compared in this Report.

7. No single document contains a complete set of internationally recognized forensic standards for hair testing. Instead, in assessing these standards, the Independent Review looked to a number of sources.

8. First, there are international guidelines on both hair analysis specifically and forensic toxicology more generally. With respect to the former, several organizations have published guidelines on hair testing for drugs of abuse and alcohol markers:

  • The Society of Hair Testing (SoHT), the only international organization dedicated to the development and standardization of hair-testing methodologies, released its first consensus statement on the examination of hair for drugs of abuse in 1997. In 2004 and 2012, the SoHT updated its consensus statement for drugs of abuse. In 2009, it released its first consensus statement for alcohol marker testing, and updated the statement in 2011 and 2014. These consensus statements contain recommended best practices on topics ranging from sample washing to testing methods.66
  • In 2001, the United Nations International Drug Control Programme released Guidelines for Testing Drugs under International Control in Hair, Sweat and Saliva (UN Guidelines) for use in national laboratories. These guidelines contained recommended procedures and analytical techniques for testing for drugs of abuse in hair.67 An updated version was released in 2014.68
  • In 2009, the German Society of Toxicological and Forensic Chemistry published Quality Requirements for the Analysis of Hair Samples, and in 2010 and 2015 the European Workplace Drug Testing Society published European Guidelines for Workplace Drug and Alcohol Testing in Hair. Both publications contain guidelines on the forensic analysis of hair for drugs of abuse and alcohol.

9. Although the above guidelines contain recommended practices for hair analysis, they do not contain all the internationally recognized forensic standards the Independent Review considered. They do not, for example, cover the general standards of practice that apply to forensic laboratories (including, for example, the need for documented standard operating procedures and the maintenance of chain of custody of samples). For those standards, the Independent Review considered published guidelines that set out best practices for forensic toxicology laboratories – guidelines not specific to hair testing. These documents include the Forensic Toxicology Laboratory Guidelines published by the Society of Forensic Toxicologists and the American Academy of Forensic Sciences (SOFT / AAFS Guidelines), as well as the version published by the United Kingdom and Ireland Association of Forensic Toxicologists.69

10. Second, internationally recognized forensic standards can be gleaned from the standards required to obtain accreditation as a forensic toxicology laboratory. These standards include those established by the Standards Council of Canada and the American Society of Crime Laboratory Directors / Laboratory Accreditation Board – organizations that accredit forensic laboratories in North America.70

11. Third, beyond what is written in the guidelines and standards, there are generally accepted “best practices” that forensic toxicology laboratories around the world have come to recognize. These practices may vary by jurisdiction but do not tend to deviate significantly from one another. In assessing the practices that were accepted internationally by forensic toxicology laboratories during the period under review, I relied primarily on the knowledge, experience, and advice of the Independent Review’s world-renowned experts, Dr. Gail Audrey Ann Cooper and Professor Olaf H. Drummer.

12. Finally, to interpret hair test results for use in forensic cases, experts rely on their own experience in the field as well as the information that can be gleaned from the academic literature. Because hair analysis is a relatively new field, the knowledge surrounding the science of hair testing continues to evolve, and hair-testing experts are expected to stay current with new research studies as they are published. Experts interpreting tests for forensic purposes are also expected to be trained in how to provide objective and impartial opinions about the results.

3. Drugs of Abuse: January 2005–August 2010

13. MDTL fell below the applicable forensic standards for hair testing for drugs of abuse from January 2005 to August 2010 in numerous ways. Below, I summarize my findings with respect to MDTL’s most obvious deficiencies.

3.1 Improper Use of ELISA Results

14. From January 2005 to August 2010, MDTL tested all hair samples for drugs of abuse using ELISA. The particular ELISA test that MDTL used came in the form of a “kit” supplied by Immunalysis Corporation (Immunalysis), a well-known and well-respected company in California. MDTL used the following ELISA kits from Immunalysis: cocaine, benzoylecgonine (a metabolite of cocaine), opiates, oxycodone, methadone, meperidine, amphetamine, methamphetamine, cannabis / THC, benzodiazepines, barbiturates, PCP, and LSD.71 The drugs and metabolites that MDTL tested most frequently during this period were cocaine, benzoylecgonine, and cannabis / THC.

15. ELISA is widely used in both clinical and forensic toxicology as a screening test. It is intended to determine quickly if a sample is negative and merits no further testing or if it is a preliminary positive, in which case the sample must be tested using another method (a confirmation test) to determine if the sample is in fact positive. The requirement to carry out a confirmation test on any preliminary positive results from immunoassay-based screens, such as ELISA, was highlighted in both the literature and in all internationally recognized hair-testing standards well before 2005.72

16. Moreover, the Immunalysis kits that MDTL used throughout the period from January 2005 to August 2010 included an insert that warned the user about the preliminary nature of the ELISA results (the example below is taken from the ELISA kit for cocaine):

THE IMMUNALYSIS COCAINE DIRECT ELISA KIT IS INTENDED FOR FORENSIC USE ONLY.

The Immunalysis COCAINE Direct ELISA Kit provides only a preliminary analyti­cal test result. A more specific alternate chemical method must be used in order to obtain a confirmed analytical result. Gas chromatography / mass spectrometry (GS-MS) is the preferred confirmatory method. Professional judgement should be applied to any drug of abuse test result, particularly when preliminary positive results are used. [Emphasis added.]

17. Despite the international consensus and the unambiguous instruction from the manufacturer about the manner in which the ELISA tests could be used, MDTL did not have the capability to test hair samples using a confirmation method, such as GC-MS or LC-MS/MS, in-house. Over the same five-year period, it also did not send all its preliminary positive ELISA samples to its reference laboratory, USDTL, for analysis.73 In fact, MDTL very rarely referred its preliminary positive samples for confirmation testing.

18. By not sending all its ELISA positive samples for GC-MS confirmation, MDTL was not in compliance with internationally recognized forensic standards (or, indeed, the instructions on the face of the ELISA kits it used). Nevertheless, from January 2005 to August 2010, MDTL reported its ELISA results – both qualitatively (as positive vs. negative) and quantitatively (in the form of a numerical drug concentration for positive samples) – to its customers. Such a practice was unacceptable in two important respects.

19. First, unlike confirmation techniques such as GC-MS or LC-MS/MS, ELISA does not and cannot identify the substances within a sample. Because ELISA relies on reactions with biological compounds (antibodies), and because other structurally related or non-drug–related compounds may bind to (or “cross-react” with) the antibodies in the test, there is a material risk of false positive results.74 Accordingly, until a confirmation test has been performed and has identified the presence of a drug or metabolite in a sample, the ELISA result cannot be reported to the customer as a positive. At best, it can be reported as a preliminary, unconfirmed, or tentative positive that requires further testing. Because MDTL did not provide such a caveat, its unconfirmed ELISA results should not have been relied on at all.

20. Second, MDTL’s practice of using ELISA to quantify drug concentrations is simply unheard of in forensic toxicology laboratories.75 Forensic toxicology laboratories use ELISA as a screening tool, meaning that it is used only as a qualitative test – to distinguish between preliminary positive and negative samples. ELISA is not designed to, and cannot, quantify drug concentrations for several reasons:

  • ELISA produces different responses for the different structurally related compounds with which it reacts or cross-reacts. For example, the ELISA kit for opiates would trigger different responses for morphine, codeine, and 6-AM (the primary metabolite of heroin). However, if a response is triggered, there is no way of knowing (without a confirmation test) whether the response means that the sample contains morphine, codeine, or 6-AM, some mixture of the three, or the presence of other metabolites or structurally related compounds.
  • Non-drug–related compounds, including substances that are naturally occurring in the body, may cross-react or interfere with the ELISA test. In general, cross-­reactivity increases the response, whereas interference decreases the response. However, there is no way of knowing whether there has been cross-reactivity or interference in a particular case.
  • The ELISA test does not create a linear response, particularly at low and high concentrations. This lack of linearity means, for example, that at higher concentrations, doubling the concentration of a drug in a sample would not result in double the response from the ELISA test. In addition, some drugs do not respond to ELISA in a linear manner at all (and for those drugs, doubling the concentration may result in up to a 10-fold increase in the ELISA response).

21. In short, the response generated by the ELISA test does not reflect, and cannot be equated with, the actual drug or metabolite concentration in the sample. A confirmation test such as GC-MS or LC-MS/MS must be carried out both to identify and to quantify the amount of the drug or metabolite present. Indeed, the warning label on the ELISA kits that MDTL used said as much.

22. Finally, even in the rare cases in which MDTL sent a preliminary positive sample to its reference laboratory for confirmation, MDTL’s practices undermined the reliability of the reported result. Not only did MDTL not advise the customer that the confirmation test had been carried out by another laboratory but it reported the results inconsistently. Joey Gareri, MDTL’s laboratory manager from May 2005 on, explained that, because MDTL used concentration ranges that were created using its ELISA results and wanted the comparison to be appropriate, the Laboratory reported its original ELISA result (with a footnote in the results report indicating it was “GC-MS / LC-MS/MS confirmed”) rather than the actual GC-MS result received from its reference laboratory. By contrast, Dr. Gideon Koren, MDTL’s laboratory director, stated that, when there was a GC-MS or LC-MS/MS result for a particular sample, the confirmed result was reported to the customer.

23. A review of the case files that the Independent Review obtained showed that MDTL’s practice was as inconsistent as Mr. Gareri’s and Dr. Koren’s recollections. In some cases, MDTL reported the reference laboratory’s result. In other cases, it reported its own ELISA result notwithstanding the fact that it had obtained GC-MS confirmation for the sample. Accordingly, even in those cases where MDTL sent a preliminary positive ELISA sample to its reference laboratory for confirmation, its reporting practices undermined the reliability of the result that was reported to the customer.

3.1.1 ELISA vs. GC-MS or LC-MS/MS Comparisons

24. MDTL began using ELISA to test for drugs of abuse in 1999 or 2000.76 Julia Klein, MDTL’s laboratory manager from June 1995 to April 2005, who was instrumental in the development of MDTL’s testing methodologies, told the Independent Review that MDTL validated its ELISA method for cocaine and benzoylecgonine, and later for opiates and cannabis / THC, by testing samples using ELISA and then sending those same samples to Psychemedics Corporation (Psychemedics), a laboratory in the United States, for GC-MS analysis. According to Ms. Klein, MDTL considered the ELISA method to be valid and appropriate because this cross-testing did not reveal any false positives, and “[a]ll cocaine positive samples which were quantitated with this method at MDTL correlated very well with the reference lab’s.” When the Independent Review asked Ms. Klein what it meant to “correlate very well” in this context, she responded that the ELISA and GC-MS results were “in the same ballpark range.”77

25. MDTL could not locate any data to support Ms. Klein’s statements about the validation of MDTL’s ELISA tests.78 As a result, the Independent Review has been unable to analyze the data on which the Laboratory apparently relied to support the validation of its ELISA method for testing for drugs of abuse. MDTL has, however, provided the Independent Review with two later sets of data comparing the Laboratory’s ELISA test results for cocaine and benzoylecgonine with the results of a GC-MS or LC-MS/MS test for the same sample:

  • The first set was data from 164 samples that were tested from June to August 2010 for cocaine and benzoylecgonine at MDTL using ELISA, and then tested at MDTL using GC-MS. MDTL compiled these data in the lead-up to the change in its testing methodology, which came into effect on September 1, 2010.
  • The second set was data from 53 samples that were tested in 2009 and 2010 for cocaine and benzoylecgonine at MDTL using ELISA, and at USDTL using LC-MS/MS. MDTL advised the Independent Review that it compiled this second set of data in or around 2012 for the purposes of the Broomfield appeal.79

26. The comparison data that MDTL provided highlights the danger of using the ELISA test in the way MDTL did for two reasons.80 First, the ELISA test generated a material number of false positive and false negative results:81

  • The false positive rates for cocaine and benzoylecgonine from the 164-sample data set were 6.7 percent and 3.7 percent, respectively, and from the 53-sample data set were 3.8 percent and 9.4 percent, respectively.
  • The false negative rates were even larger for the 164-sample data set, at 15.3 percent and 19.1 percent for cocaine and benzoylecgonine, respectively. Although there were fewer false negatives in the 53-sample data set (no false negatives for cocaine and 2.3 percent false negatives for benzoylecgonine), this decline was likely because there were fewer negative samples in total in that data set.
  • Overall, more than 22 percent of the samples tested in the 164-sample data set and more than 3 percent in the 53-sample data set had either a false positive or a false negative response for either cocaine or benzoylecgonine.

27. Second, the ELISA test also did not perform well quantitatively:82

  • The average GC-MS result for cocaine in the 164-sample data set was almost half the average ELISA result, and the median GC-MS result was less than half the median ELISA result. In addition, for cocaine, 10 samples showed a difference between the GC-MS and the ELISA result of over 300 percent.
  • However, the opposite was true for benzoylecgonine: the average GC-MS result for benzoylecgonine in the 164-sample data set was more than double the average ELISA result, and the median GC-MS result was slightly less than double the median ELISA result.

28. These comparison data confirm that MDTL’s preliminary and unconfirmed ELISA test results were unreliable, both qualitatively and quantitatively, for forensic purposes.

3.2 Absence of Standard Operating Procedures and Contemporaneous Documentation

29. MDTL has been unable to produce any written analytical procedures that were in use for its drug tests from 2005 to 2010. Ms. Klein told the Independent Review that MDTL created written procedures in the late 1990s or early 2000s for the tests it performed and amended these procedures over the years. Although MDTL provided the Independent Review with copies of procedures it located on Ms. Klein’s old computer, it explained that the origin, use, and other details of those documents (which were undated) were unknown and, most importantly, did not reflect the analytical processes that were in place from 2005 to 2010. Tatyana Karaskov, MDTL’s laboratory technician who performed virtually all the ELISA tests during this period, also recalled that she had once prepared a written list of the analytical steps taken in carrying out the hair tests for drugs of abuse. However, MDTL was unable to produce the list that Ms. Karaskov described for the Independent Review.

30. The only written procedure that the Hospital for Sick Children (SickKids or the Hospital) produced from the 2005–10 period for drugs of abuse was a document entitled “Motherisk Drug Testing Laboratory Procedures.” This document was prepared by MDTL in conjunction with SickKids’s research finance department to give guidance to staff involved in the receipt of samples and the reporting of results.83 These procedures were launched in July 2007 in conjunction with MDTL’s new electronic reporting system. However, all the procedures detailed within the document were non-analytical, dealing with the numbering and logging (referred to as accessioning) of samples and the invoicing of customers. The document did not provide guidance to staff on any of the analytical methods used by MDTL, including how to prepare hair samples for analysis, how to carry out the tests, how to assess whether the tests were acceptable, or how to interpret the findings.

31. In addition, from 2005 to 2010 MDTL did not maintain contemporaneous documentation of the steps it performed for a specific sample, including records of who performed which step in the process, what was performed, or when or how it was performed.

32. It was not until the fall of 2010, after MDTL had changed its testing methodology for drugs of abuse (which came into effect on September 1, 2010), that the Laboratory adopted formal standard operating procedures for its hair analysis for drugs of abuse. At that time, it also began to keep a drug-testing log for each sample to record the preparation steps that were taken. MDTL’s post-2010 procedures are described below.

33. Standard operating procedures form an integral part of any quality management system. They provide laboratory staff with a clear description of the process to be followed for a particular test, from sample preparation through to the reporting of the result – in other words, a step-by-step guide of how to carry out the analytical test. In this way, standard operating procedures help to ensure that laboratory staff have a fixed process to follow for the analytical tests carried out by the laboratory as well as the non-analytical aspects of laboratory practice (including, for example, safety and document retention). Standard operating procedures should be kept up to date, and staff should be instructed not to deviate from them.

34. From 2005 to 2010, it was standard practice for both clinical and forensic laboratories to have standard operating procedures in place for each of the analytical tests they performed. In addition, forensic laboratories were required to maintain records of the tests they performed, including documents identifying what steps were taken in respect of a sample, who took them, when and why, as well as any decisions that were made that might be relevant to the result. In the forensic context, standard operating procedures, together with contemporaneous documentation of the steps taken, have the added benefit of maintaining the chain of custody of the sample within the laboratory.

35. Although the absence of documented procedures or contemporaneous documentation does not automatically mean that a laboratory’s testing methodology is unreliable, it does raise doubts about how a particular test was carried out. Without standard operating procedures, questions may arise about how the samples were handled, whether the tests were carried out correctly, how the laboratory staff ensured that samples were not contaminated, when mistakes occurred, and whether any corrective actions were discussed and implemented by the laboratory’s management.

36. Although MDTL did not have any standard operating procedures for its analytical methods from 2005 to August 2010, the Laboratory told the Independent Review that, to a large extent, the standard operating procedures that were implemented in December 2010 codified the existing practices at the time (with the exception of the testing methodology used). For the purposes of the Independent Review, Ms. Karaskov also prepared a summary containing her best recollection of how the hair samples were handled and how the ELISA test was carried out before 2010. However, MDTL acknowledged that its processes may have evolved over time and its practices may have varied in different ways from what was contained in either the December 2010 procedures or Ms. Karaskov’s summary.

37. Without any standard operating procedures or contemporaneous documentation for samples tested from 2005 to August 2010, neither the December 2010 procedures nor Ms. Karaskov’s summary provides a reliable method of determining the specific steps taken in a particular case. In fact, the Independent Review identified material inconsistencies in the manner in which MDTL carried out its hair tests at various stages of the analysis during this period.

3.2.1 Sample Suitability

38. Without any standard operating procedures, MDTL appeared to rely entirely on the discretion of its laboratory staff to determine whether a sample was suitable for testing. Ms. Karaskov told the Independent Review that, from 2005 to August 2010, her practice was to use up to 10 mg of hair and not less than 2.5 mg. In some cases (e.g., with neonatal samples, where only a very small sample of hair was available), she would exercise her judgment to decide whether to run the test and report the results or to advise the customer that the test result was “NSQ” (not sufficient quantity) and request a larger volume of hair.

39. However, whether a sample is suitable for testing should not depend solely on the judgment call of staff but rather on the proper validation of the laboratory’s instrumentation. There were examples in the case files reviewed where a hair sample was cut into multiple segments of very low weight, within the range that Ms. Karaskov described (as low as 3.5 to 5.2 mg each). There were also cases in which the sample was too long to be reliably tested in bulk (in one case, for example, MDTL tested a whole-length, or bulk, sample that was 42 cm long and weighed only 9.7 mg).84 Ultimately, the absence of standard operating procedures from 2005 to August 2010 for sample volume created inconsistencies in MDTL’s analytical methods that call into question the reliability of MDTL’s results.

3.2.2 Washing

40. The Independent Review received conflicting accounts about MDTL’s washing procedures from 2005 to 2010. Ms. Karaskov’s recollection was that it was her practice to wash every sample (adult, child, or neonate) after it was cut and before testing. However, Ms. Klein told the Independent Review that, during her tenure (until April 2005), not all hair samples were washed before testing and that neonatal samples were never washed. According to Ms. Klein, she and Ms. Karaskov together decided whether a sample would be washed. Similarly, Mr. Gareri recalled that MDTL’s practice before 2010 was not to wash all hair samples and to do so only when washing was specifically requested. When the Independent Review asked Dr. Koren for his understanding of MDTL’s practice for washing hair samples from 2005 to 2010, he deferred to Ms. Karaskov.

41. The issue of decontamination is further complicated by the following circumstances:

  • The December 2010 standard operating procedures (which, according to MDTL, had largely codified its 2005 to 2010 actual practices) provided that, after 2010, only adult samples were washed; neonatal and child samples were not. The standard operating procedures were, accordingly, inconsistent with Ms. Karaskov’s recollection that she washed all pre-2010 samples (including child and neonatal samples) after cutting and before testing.
  • From 2001 to 2014, MDTL participated in proficiency testing for hair testing for drugs of abuse – a test administered by the SoHT. In its proficiency test submissions, MDTL was required to submit information regarding its methods to the SoHT. In its 2003 and 2006 submissions, MDTL reported that, “routinely, not all the hair samples are washed in our lab prior to extraction. When we need to wash them, we use Tween 20 in a dilute solution (1:20).” Other than stating “washing of hair” under the heading “clean-up” in the 2001 proficiency test submission, none of MDTL’s other submissions to the SoHT before 2010 contained any reference to hair washing either way.
  • MDTL’s new electronic reporting system, which came into effect in July 2007, gave staff the option to record if samples were “washed” or “unwashed.” However, in the case files that the Independent Review obtained, with one exception the only reference to hair samples being “washed” on the results report was in relation to samples that were being tested for FAEEs.85 Despite the ability for staff to indicate “washed” or “unwashed” on MDTL’s results reports after July 2007, the reports were silent on whether the sample was actually “washed” or “unwashed” before testing for drugs of abuse. There was also no mention in the interpretation reports that the Independent Review obtained from 2005 to 2010 of whether the hair samples were washed.

42. Washing a hair sample before analysis is a step that can significantly affect the result. However, without any written procedures or documentation regarding a step as critical as the washing of hair samples, there can be no level of certainty about the process that MDTL followed before 2010. At best, MDTL washed hair samples when requested to do so by the customer or when the laboratory technician who was preparing the sample decided to wash the sample. Because no record was kept of whether a sample was washed or even which laboratory technician prepared the sample for testing, there is no reliable way before September 2010 to identify whether a particular hair sample was washed. In fact, in view of the information set out above, it is more likely that samples (adult, child, and neonate) were not routinely washed before analysis during this time.

3.2.3 Extraction Method

43. Without standard operating procedures, there also appeared to be inconsistencies in MDTL’s extraction methods. Table 5.1 summarizes the extraction methods that MDTL reported to the SoHT in its proficiency test submissions from 2001 to 2010.

Table 5.1 Summary of extraction methods reported for ELISA by MDTL to the SoHT, 2001–10

Year

Solvent Extraction

2001

MeOH, sonicate 30 mins, 45°C for 18 hrs

2002–3

MeOH, sonicate 60 mins, 45°C overnight

2004–5

MeOH, sonicate 30 mins, 56°C for 18 hrs

2006

MeOH, sonicate 60 mins, 52°C overnight

2007

MeOH, sonicate 30 mins, 52–56°C for 18 hrs

2008

MeOH, sonicate 30–60 mins, 56°C overnight

2009

MeOH, 56°C overnight (18 hrs)

2010

MeOH, 56°C (~16 hrs)

44. As set out in Table 5.1, the reported solvent extraction method used by MDTL to release the drugs from the hair matrix varied over the period from 2001 to 2010. Although it would not be uncommon for methods to vary over time with improvements or changes in instrumentation, times and temperatures in the extraction process that appear to have changed from one year to the next sometimes reverted back to a previous method.

45. Given the impact that extraction procedures may have on the results generated, it is critical for both forensic and clinical laboratories to maintain standard extraction procedures or, at the very least, to keep a record of the extraction methods employed in a particular case.86 Laboratories must investigate their extraction methods, and, if asked, they should be able to justify and explain any changes in methodology. When asked, however, MDTL was unable to produce any data to demonstrate the steps taken to validate the analytical methods used from 2005 to August 2010, including its extraction procedures. Accordingly, the Independent Review’s experts were unable to assess the impact of the varying extraction methods on the results that MDTL generated.

3.2.4 Calibrators

46. Without standard operating procedures, MDTL’s use of calibrators was also inconsistent. Spreadsheets containing MDTL’s ELISA data revealed that the number of calibrators that MDTL used with a batch often varied from four to six. In addition, there were case files that contained multiple spreadsheets (suggesting that the sample was tested more than once) where the optical densities (absorbances) generated by the ELISA instrument remained the same for the sample, but the optical densities for the calibrators changed. Because the calibration curve changed, the test results for the sample also changed (even though there was no change in the underlying data). These data indicate, at least in one of the spreadsheets, that the calibrators were run in a separate batch from the actual sample. If that is, in fact, what happened, such a practice would be entirely unacceptable.87 In any event, MDTL’s lack of standard operating procedures and contemporaneous documentation means that, for those cases, there is no explanation for why the tests were carried out in this manner.

3.2.5 Quality Controls

47. MDTL was also inconsistent in how it used quality controls: Although positive and negative controls were run in many cases, they were not run with every batch. In the batches in which no quality controls were run, there would have been no independent measure to assess if the batch had been run properly.

48. MDTL has been unable to explain why quality controls appear to have been run in some batches but not in others. Moreover, even for the batches in which quality controls were run, MDTL did not have any processes in place to monitor performance of the quality controls from batch to batch. Such batch-to-batch monitoring is essential for ensuring the reliability of the Laboratory’s analytical processes and results, and in the absence of such monitoring, the Independent Review was unable to assess when batches failed and why.

3.2.6 Criteria for Acceptability

49. Many of MDTL’s case files contained multiple spreadsheets for the same sample, suggesting that the samples were tested multiple times. Although it is not uncommon to rerun a batch (e.g., if one of the quality controls fails or if there is some malfunction in the instrument), MDTL did not have any written procedures in place that contained the criteria to be applied in deciding whether to accept or reject a particular batch.

50. Ms. Karaskov told the Independent Review that she would retest a batch if she suspected there was a malfunction in the instrument. In addition, if a particular sample had a concentration greater than the highest known concentration used to create the calibration curve, she would dilute the sample, retest the diluted sample, and then correct the result using a dilution factor (a flawed practice).88 In any event, neither the spreadsheets nor the case files that MDTL obtained explained why additional ELISA tests were carried out – whether it was as a result of a failed quality control, instrument malfunction, the use of the (flawed) dilution technique that Ms. Karaskov described, or something else – and why the results of one spreadsheet were selected for reporting over the others.

51. Ultimately, MDTL may well have had legitimate reasons to retest a particular sample or to report its results in the manner in which it did. However, the absence of documentation to back up its decision making calls into question the reliability of the results that MDTL reported.

3.2.7 Cut-offs

52. Although there were no written procedures establishing the cut-offs that MDTL was using from 2005 to August 2010, MDTL told the Independent Review that it used a 0.10 ng/mg cut-off for all drugs and metabolites during this period (until it changed the cut-off for cocaine, oxycodone, methadone, and meperidine to 0.2 ng/mg in 2009).

53. In addition to having a cut-off to distinguish between positive and negative samples, MDTL also used the term “trace” as an in-between category. There was no defined range for MDTL’s reporting of trace results from its ELISA test. According to MDTL, trace referred to a sample that was “clearly non-negative based on the optical density relative to the blank but did not quite reach the lowest calibrators included on the curve.” In forensic toxicology laboratories, such a result would normally be referred to as a “borderline negative” or “borderline preliminary positive” and sent for confirmation testing. At MDTL, however, the result was reported as trace without further confirmation.

54. MDTL’s use of the term “trace” in the context of an unconfirmed ELISA result was inappropriate. The term is most commonly associated with a confirmed test result following GC-MS or LC-MS/MS analysis where the presence of the drug has been confirmed but is at a concentration that cannot be accurately measured (i.e., below the limit of quantification but above the limit of detection of the instrument). However, because the ELISA test cannot identify the presence of a drug or metabolite, MDTL should not have used the term at all to refer to its ELISA results.

55. In any event, MDTL appeared also to rely on inconsistent cut-offs from 2005 to 2010. The cut-offs that MDTL gave to the Independent Review did not match the ones that it had previously reported to the SoHT in its proficiency test submissions during this period (see Table 5.2).

Table 5.2 Cut-offs reported to the SoHT in MDTL’s proficiency test submissions, 2005–10

Drug

2005

2006

2007

2008

2009

2010

Cocaine

0.2

0.2

0.2

0.2

0.2

Benzoylecgonine

Opiates

0.1

0.2

0.1

0.1

0.2

Amphetamine

0.2

0.2

0.2

0.2

0.2

Methamphetamine

0.2

0.2

0.2

0.2

0.2

Cannabis/THC

0.1

0.1

0.15

0.15

0.15

0.2

PCP

Barbiturates

Benzodiazepines

0.2

0.2

Note: Dashes indicate that no cut-off was reported on the SoHT submission form.

56. As set out in Table 5.2, many of the cut-offs that MDTL provided to the SoHT were greater than the 0.1 ng/mg that MDTL reported using for the vast majority of its tests from 2005 to 2010. Although Mr. Gareri told the Independent Review that MDTL sometimes used higher cut-offs if the ELISA kits were not working well at the low end of the calibration range, it was not clear from the data and the case files that were provided to the Independent Review when a different cut-off was used.

57. Moreover, MDTL’s case files contained numerous examples of results that were inconsistent with MDTL’s reported cut-offs and with each other. For example, in one case a result of 0.05 ng/mg for benzoylecgonine was reported as trace, but in another case a result of 0.09 ng/mg for benzoylecgonine was reported as negative. Similarly, in one case a result of 0.01 ng/mg for cannabis / THC was reported as trace, whereas a result of 0.03 ng/mg for cannabis / THC in another case was reported as negative.

58. Since cut-offs are what distinguish a positive result from a negative one, the propriety of using a particular cut-off for a particular test needs to be assessed through validation of the laboratory’s procedures. In addition, cut-offs need to be fixed by the laboratory in documented procedures, to ensure that staff are aware of the limits of the instruments being used as well as how and when the results of the test should be reported. Again, MDTL’s lack of standard operating procedures from January 2005 to August 2010, as well as the inconsistencies that the Independent Review identified, undermines the reliability of the results that it reported during this period.

3.3 Absence of Oversight

59. From 2005 to 2010, virtually all the ELISA tests for drugs of abuse were carried out by Ms. Karaskov. She performed the test, read the data from the ELISA instrument, calculated the drug or metabolite concentration from that data, and ultimately reported the result. As laboratory director, Dr. Koren signed each results report before it was released. However, it is not clear why Dr. Koren signed the form because, from at least May 2005 (when Mr. Gareri replaced Ms. Klein as laboratory manager) to August 2010, neither Dr. Koren nor anyone else within MDTL reviewed or checked Ms. Karaskov’s work.89

60. Dr. Koren told the Independent Review that, from his perspective, because he was not trained as a laboratory technician, when he signed the results reports he was “not verifying himself that the actual numbers were correct”; rather, “he was signing to confirm that he had confidence in the technician who performed the study (Karaskov).” In this way, rather than implement any oversight mechanisms within the Laboratory, MDTL appeared to rely on Ms. Karaskov’s many years of experience and judgment to ensure that the tests were being run properly and the results recorded accurately.

61. For his part, Mr. Gareri was not charged with overseeing – and lacked the expertise to oversee – Ms. Karaskov’s work, describing it to the Independent Review as a “black box” to him. Over time, however, he became increasingly concerned about the fact that Ms. Karaskov was carrying out virtually all the ELISA tests for drugs of abuse (an average of 2,000 samples per year) without any form of oversight. He told the Independent Review that his concern did not stem from any particular unease with the quality of Ms. Karaskov’s work or any lack of confidence in her abilities. Rather, Mr. Gareri became concerned that “no one could be mistake-free given the volume of samples that [Ms. Karaskov] was handling,” and he considered the fact that she alone was responsible for performing all the ELISA tests to be a “weakness for the lab.”

62. Ultimately, clinical and forensic laboratories do not put quality assurance and oversight mechanisms in place because they lack confidence in their staff. Oversight is required to ensure that the laboratory’s analytical methods are in compliance with the recognized standards of practice and that its tests are performed properly each time.

63. In fact, Mr. Gareri’s concerns appear to be borne out in MDTL’s case files. Most of the case files that the Independent Review obtained included copies of the spreadsheets that were used to calculate the ELISA results. Although the vast majority of the results on the spreadsheets matched the results found in the results reports provided to customers, in a number of cases the reported result did not match the result contained in the spreadsheets. These discrepancies were most common for cocaine and benzoylecgonine results; however, the Independent Review also found cases in which the results reports did not match the spreadsheets for opiates, benzodiazepines, meperidine, and cannabis / THC.90

64. The Independent Review identified three types of reporting discrepancies. First, there appeared to be simple transcription errors. For example, in one case the spreadsheet generated a result of 3.83 ng/mg for cocaine, but the result was reported as 3.3 ng/mg. Other examples included a spreadsheet that generated a result of 0.08 ng/mg for benzoylecgonine, but was reported as 0.10 ng/mg, and a meperidine result of 1.4 ng/mg that was reported as 1.2 ng/mg.

65. Transcription errors are natural when laboratory staff are tasked with manually inputting results from one source to another, particularly where there is a large amount of data to input. Accordingly, it is good laboratory practice to have a process in which one staff member independently checks the data transfer performed by another staff member. However, MDTL did not have such a practice in place from 2005 to August 2010.

66. Second, and of greater concern, were the few cases in which the results generated for cocaine and benzoylecgonine appeared to have been transposed. For example, in one case the spreadsheets generated results for cocaine and benzoylecgonine of 0.9 and 1.51 ng/mg, respectively; however, the results report provided to the customer gave the result for cocaine as 1.51 ng/mg and for benzoylecgonine as 0.9 ng/mg. MDTL did not provide any explanation for this apparent transposition of the cocaine and benzoylecgonine test results.

67. Third, in a few cases the reported results did not appear anywhere in the spreadsheets in the case file and the discrepancies did not appear, at least on their face, to result from a transcription error. For example, as set out in Table 5.3, in one case in which the sample was tested in three segments, the results for benzoylecgonine were properly recorded in the results report, but the results for cocaine were not contained in any of the spreadsheets provided and, in fact, varied significantly from the results generated in the spreadsheets.

Table 5.3 Reported results and results from spreadsheets for cocaine and beonzoylecgonine in sample case (3 segments)

Cocaine

Benzoylecgonine

Reported result

Spreadsheet result

Reported result

Spreadsheet result

Segment 1

0.95

2.33

Trace

0.09

Segment 2

1.98

5.14

0.18

0.18

Segment 3

3.30

10.27

0.29

0.29

68. MDTL located the spreadsheets used to generate the results in the Broomfield case and produced them to the Independent Review. There were three spreadsheets for each of the cocaine and benzoylecgonine tests for the child’s hair sample, which suggests that the sample was tested three times for each of cocaine and benzoylecgonine.91 MDTL was unable to explain why the tests appear to have been run in triplicate. Moreover, the results found in the six Broomfield spreadsheets were startling (see Table 5.4).

Table 5.4 Reported results and results from spreadsheets for cocaine and benzoylecgonine in child’s sample collected August 9, 2005 (15 segments)

Cocaine

Benzoylecgonine

Reported result

Spread- sheet #1

Spread-sheet #2

Spread- sheet #3

Reported result

Spread- sheet #1

Spread- sheet #2

Spread- sheet #3

Seg 1

7.56

37.51

1.90

49.41

 

1.90

7.56

3.76

5.93

Seg 2

6.22

34.26

4.40

51.39

 

4.40

6.22

3.93

7.60

Seg 3

11.66

53.81

11.45

35.17

 

11.45

11.66

5.9

11.64

Seg 4

25.43

137.17

16.41

1582.18

 

16.41

25.43

16.39

33.97

Seg 5

17.81

157.01

11.95

614.75

 

11.95

17.81

29.36

83.20

Seg 6

62.04

466.94

62.04

615.23

 

24.27

24.27

60.44

213.60

Seg 7

75.88

644.76

75.88

341.53

 

35.99

35.99

86.22

336.60

Seg 8

44.31

436.17

44.31

341.53

 

34.56

34.56

60.42

427.32

Seg 9

52.85

196.46

52.85

284.07

 

32.06

38.75

62.97

170.52

Seg 10

22.64

377.97

20.36

467.46

 

20.36

29.51

39.57

143.42

Seg 11

28.14

368.27

28.14

234.49

 

14.30

37.23

32.31

71.44

Seg 12

31.89

324.22

31.89

548.86

 

11.46

32.06

25.85

55.09

Seg 13

30.80

191.70

30.80

486.51

 

14.30

22.64

15.54

25.95

Seg 14

11.45

116.99

4.65

264.28

 

4.65

14.3

10.17

18.80

Seg 15

7.66

45.58

7.66

171.10

 

4.16

11.46

6.12

14.87

69. The spreadsheets revealed significant discrepancies between the spreadsheets and the results reported. As set out in Table 5.4, for segments 1 to 5, the cocaine and benzoylecgonine results appeared to have been transposed (purple and yellow). For segments 6 to 8, the results appeared to have been accurately transcribed because they matched at least one of the spreadsheets for both cocaine and benzoylecgonine (purple and yellow). For segments 9, 11 to 13, and 15, the reported cocaine results appeared to have been accurately transcribed (purple), but the reported benzoylecgonine results were not found anywhere in any of the spreadsheets (blue). Finally, for segments 10 and 14, the result in the cocaine spreadsheet was reported as a benzoylecgonine result (yellow), but the reported cocaine result was nowhere in the spreadsheets (blue).

70. MDTL has been unable to explain any of the discrepancies between the reported results and the spreadsheets that were contained in the case files. Although Ms. Karaskov told the Independent Review that “tests might have been re-run” and that the Independent Review may have been missing some spreadsheets, even missing spreadsheets would not explain the manner in which the results were reported. Indeed, there is no logical reason for MDTL to report the results from one spreadsheet for certain segments and the results from another spreadsheet for others. These discrepancies, which in some cases were significant, further demonstrate the unreliability of MDTL’s reported results.

3.4 Failure to Wash Samples Routinely

71. Washing a hair sample before analysis is an important stage in the process because it helps to remove dirt and grime as well as other substances on the surface of the hair, including drugs that are present from external contamination. In fact, washing is recommended, and the issue of external contamination is required to be considered in the SoHT’s consensus statements for drugs of abuse in both 1997 and 2004.92 The SoHT also recommended that the solvent and the solution (the washings) be stored for further analysis if required.93

72. Because external contamination may be significant, there can be substantial variation between the test results for washed and unwashed hair. In its proficiency test submission to the SoHT in 2003, MDTL reported GC-MS results for hair samples that were tested twice: unwashed and washed.94 There was significant variation in MDTL’s submissions for the unwashed and washed results (see Table 5.5).

73. Accordingly, even MDTL’s own records from as early as 2003 reveal that washing can have a substantial impact on the results, in some instances decreasing the result by a factor of four.

Table 5.5 Results submitted to SoHT in MDTL’s 2003 proficiency test submissions

Sample

Unwashed

Washed

Sample A

Cocaine

1.97 ng/mg

0.54 ng/mg

Benzoylecgonine

1.31 ng/mg

0.43 ng/mg

Sample B

Cocaine

2.08 ng/mg

0.46 ng/mg

Benzoylecgonine

0.78 ng/mg

0.19 ng/mg

Sample C

Cocaine

2.10 ng/mg

0.58 ng/mg

Benzoylecgonine

2.80 ng/mg

0.69 ng/mg

Note: MDTL reported that these results were obtained by GC-MS.

74. Whether a laboratory’s practice of testing unwashed hair samples is acceptable depends on the manner in which the result is used and interpreted and on the disclosure made to the customer. Washing is mandatory for tests that are carried out to determine drug use (ingestion). In contrast, in cases where the hair sample is being tested to determine environmental exposure to a drug rather than active drug use, it may be acceptable not to wash the sample.

75. I have already concluded that MDTL likely did not wash its hair samples routinely when testing for drugs of abuse from 2005 to August 2010. Nor did MDTL routinely store any washings for later analysis. Although the MDTL leadership told the Independent Review that MDTL initially considered tests of unwashed hair samples to be acceptable, the Laboratory’s practice was inconsistent with internationally recognized forensic standards, given how MDTL’s results were used: MDTL relied on the results from its hair tests to reach conclusions about an individual’s drug use (rather than just exposure). For example, in Broomfield, the accused was charged and convicted of an offence based on the child’s alleged ingestion of cocaine over a period of 14 months. Whether the child’s hair samples were washed before testing was important to the issue of whether the child had ingested cocaine or been inadvertently exposed to it during that period.

76. In that case, Dr. Koren and Ms. Karaskov testified at the trial in 2009 that the child’s hair samples were washed before testing. However, MDTL did not produce any contemporaneous documentation to establish whether the samples were washed, when they were washed, how they were washed, and by whom. A crucial piece of the information required to interpret the results that MDTL reported is missing. If the sample in Broomfield was not washed, no conclusions could be drawn about whether the child had ingested, rather than been exposed to, cocaine.

77. Finally, even in cases where a laboratory is testing for exposure (rather than use), the laboratory must clearly explain that the hair was not washed before testing and the reasons for its decision not to do so. From 2005 to August 2010, there was no indication one way or another (with the one exception described above) about whether the sample was washed before testing for drugs of abuse. Accordingly, users of the test results would not have known or understood that the samples likely were not washed or, indeed, the significance of MDTL’s failure to do so. The reports also contained no reference to washings, including any explanation for why the washings from any samples that were washed were not analyzed.

3.5 Inadequate Chain-of-Custody Procedures

78. While many of the concepts I have described in this chapter arise in both forensic and clinical toxicology, chain of custody is a uniquely forensic principle. The purpose behind maintaining a chain of custody is to ensure, to the extent possible, that a sample can be traced from the time it was collected until the time the test is completed in order to minimize the risk of tampering or contamination along the way.

79. Although MDTL appeared to recognize the importance of maintaining the chain of custody by using Chain of Custody Requisition forms (which recorded who collected the sample and which courier company delivered it to the Laboratory, and required samples to be sealed before delivery) for samples that were collected outside the Hospital, it did not meet the accepted forensic standards in practice in several ways:

  • Many of the Chain of Custody Requisition forms in MDTL’s case files were incomplete and some were missing. Accordingly, even though MDTL had correctly identified the need to maintain chain of custody of its samples by using these forms, it did not ensure that it routinely maintained chain of custody in practice.
  • MDTL did not require Chain of Custody Requisition forms for samples collected within the Hospital. As a consequence, for those samples there was no record of whether the sample was sealed or when the sample was received within MDTL and by whom.95
  • No chain-of-custody documentation was kept at all once MDTL received the sample, whether it originated from an outside source or from within the Hospital. Indeed, as noted above, MDTL kept no records at all of who handled a particular sample, for what purpose, or when.

80. The problems associated with a lack of chain of custody within the Laboratory are compounded by the fact that MDTL did not limit access to its samples only to the laboratory technicians who were preparing and testing the hair samples. Ultimately, MDTL operated just like any other research laboratory in a hospital setting. Students and researchers had unfettered access to the Laboratory, including the locations in which the samples were stored. Although visitors to the Laboratory needed a pass to enter, they were not required to sign in or out.

81. MDTL’s conclusion that there was no need to maintain the chain of custody of a sample once it was within the Hospital or the Laboratory demonstrates a misunderstanding about the purpose for maintaining chain of custody and ensuring the traceability of samples. Missing or incomplete Chain of Custody Requisition forms can raise questions about the integrity of the samples submitted for analysis. Moreover, because MDTL staff did not witness and document sample transfers, there was an increased risk of sample contamination or human error.

3.6 Inadequate Record Keeping

82. Forensic laboratories are required to maintain their data and documentation for a specific case for the purpose of a subsequent independent review of the test. Some forensic laboratory guidelines refer to this package of information as a “litigation pack.”

83. Although MDTL did not maintain adequate records of the tests it carried out, even for the records it did maintain, its practices were deficient: MDTL did not maintain its documents relating to a particular case in a central location. When the Independent Review requested case files from 2005 to 2010, MDTL staff needed to pull the documents for each file from several different locations. For example, the original customer requests and log sheets were kept separate from the data and spreadsheets used to calculate the ELISA results, which in turn were kept separate from any interpretation reports that MDTL prepared.

84. Moreover, of the case files the Independent Review received, many were incomplete and missing documents that could not be located by MDTL staff. For example, in some instances the case file contained a letter to MDTL requesting the hair test, but contained no Chain of Custody Requisition Form. In other cases, there was no record of the original request at all, making it impossible to assess what the customer had requested or to determine if MDTL had carried out the request. In those cases, it is unclear whether no written request was made or whether the request was made but has since been lost.

85. There were also a number of cases in which the spreadsheets used by MDTL to generate the ELISA result could not be found. Without this underlying data, a reviewing toxicologist would be unable to assess independently the test that MDTL performed, including to determine if controls were run and whether the result was calculated properly.

86. The problems that can be created by poor record-keeping practices are demonstrated by what occurred on the Broomfield appeal. On this appeal, the Crown and the defence requested that MDTL produce all notes and records relating to the samples tested.96 On August 11, 2013, in response to a production request from the defence, Mr. Gareri stated:

Any laboratory records that existed pertaining to the analysis of [the child’s] hair have already been submitted to the Police and/or the Crown. There are no additional lab notes or other records available.

87. Mr. Gareri also noted that records relating to the child’s hair samples might be unavailable because of a ceiling leak in his office and because a computer that was used to record laboratory information had died.

88. After the Independent Review was initiated, MDTL did locate the spreadsheets for the child’s samples as well as records relating to a referral of the accused’s hair sample to USDTL.97 MDTL explained that it was able to locate the spreadsheets because it had found a backup of Ms. Karaskov’s computer.

89. The problem, however, was that the appeal was heard and decided without the benefit of the spreadsheets containing the ELISA results for the child’s samples. The spreadsheets were not made available to the Crown, the defence, or the experts who were retained on the appeal. MDTL should have had a better system of maintaining its records, to ensure that any records in its possession with respect to a forensic case were readily available and retrievable. It is unacceptable that records as important as the cocaine and benzoylecgonine spreadsheets were not made available to the Crown and to the defence on the Broomfield appeal.

4. Drugs of Abuse: September 2010–April 2015

90. In the fall of 2010, MDTL made two significant changes in its operations: The Laboratory transitioned to routine GC-MS confirmation for the vast majority of the drugs and metabolites it tested for in adult and child samples, and it implemented standard operating procedures as part of a quality management system. Despite the improvements, however, several significant deficiencies remained in the manner in which MDTL carried out its hair analysis from September 2010 to April 2015. These deficiencies call into question the adequacy and reliability of MDTL’s results during this period.

4.1 Improvements in MDTL’s Analytical Practices

91. Many of the flaws in MDTL’s analytical methods that I identified from January 2005 to August 2010 were corrected in the fall of 2010.

4.1.1 GC-MS Confirmation and Elimination of Quantitative ELISA Results

92. Beginning on September 1, 2010, MDTL began routinely to confirm its preliminary positive ELISA results for the vast majority of drugs and metabolites tested in adult and child samples. MDTL had developed a GC-MS method that tested for 17 drugs and metabolites simultaneously – cocaine, benzoylecgonine, norcocaine, cocaethylene, 6-AM, morphine, codeine, hydrocodone, hydromorphone, oxymorphone, oxycodone, methadone, meperidine, amphetamine, methamphetamine, MDMA, and MDA. When an ELISA screen produced a positive response for any drug or metabolite,98 provided there was sufficient sample to carry out a second analysis, MDTL would perform a confirmation test using GC-MS in which it tested for all the drugs and metabolites at the same time.

93. Given that a positive result from an ELISA screen led directly to a GC-MS confirmation test, a screening test was not necessarily used for every positive result that MDTL ultimately reported. For example, if the ELISA test provided a preliminary positive result for cocaine, the sample would be sent for GC-MS confirmation. The GC-MS test might also detect morphine in the sample. In that instance, MDTL would report the positive morphine result to the customer even though no ELISA opiates screen had ever been used. In cases in which an ELISA screen was positive but there was not enough hair to test the sample using GC-MS, MDTL reported the ELISA result qualitatively (as positive or negative) and included the following note on the results report: “Non-sufficient Quantity of hair received for GC-MS confirmation. Recommend submission of new sample.”

94. There were, however, five drugs or classes of drugs for which MDTL had not developed a GC-MS test for adult and child samples: benzodiazepines, barbiturates, PCP (phencyclidine, commonly known as angel dust), LSD, and cannabis / THC. Because no in-house GC-MS method was available, MDTL continued to use ELISA only (except in cases where it sent the samples to its reference laboratory for confirmation). However, MDTL stopped calculating the quantitative value and reported only whether the result was positive (above the cut-off), negative, or “trace.”

95. MDTL’s procedure differed for neonatal hair samples, which the Laboratory defined as being samples from infants under six months old. Because MDTL could not develop a method on its GC-MS instrument that it considered sufficiently sensitive to test neonatal samples (frequently, only a lower amount of sample is available), it continued to test these samples with ELISA for all drugs and metabolites. However, as with the adult and child samples, it reported only the qualitative result. In some (but not all) cases, where MDTL used only ELISA, the Laboratory referred the sample to its reference laboratory for confirmation testing.

96. In 2014, MDTL underwent another change in its testing methods when it stopped using GC-MS analysis to test for drugs of abuse. Starting on May 10, 2014, MDTL used LC-MS/MS to test simultaneously for 18 drugs and metabolites for all samples, including for neonates (the same 17 drugs and metabolites that it tested using GC-MS, plus cannabis / THC), without first screening the sample using ELISA. Because MDTL did not use ELISA as a preliminary screening test for those 18 drugs and metabolites, after May 2014 ELISA was used only to test for the four drugs or classes of drugs for which MDTL had not developed any GC-MS or LC-MS/MS method: benzodiazepines, barbiturates, PCP, and LSD. As before, MDTL did not report the quantitative value generated by the ELISA test for those drugs but only whether the result was positive or negative.99 The drugs and metabolites that MDTL tested and the testing methods used from September 2010 to April 2015 for adult and child samples are found in Table 5.6, and for neonatal samples in Table 5.7.

97. Three changes that MDTL made to its analytical testing methods in September 2010 significantly improved the adequacy and reliability of its results. First, MDTL’s move from using ELISA only to routine GC-MS confirmation for adult and child samples was a dramatic improvement. All internationally recognized hair-testing standards required confirmation of immunoassay-based screening test results by a confirmation test such as GC-MS. By routinely confirming its ELISA preliminary positive results with GC-MS for 17 drugs and metabolites in adult and child samples, MDTL was finally using the appropriate instrument to carry out those tests.100 MDTL’s move to LC-MS/MS analysis for adult, child, and neonatal samples in May 2014 represented a further step in improving the adequacy and reliability of its analytical results.101

98. Second, MDTL’s decision to stop reporting quantitative results from ELISA for the drugs and metabolites for which it did not have GC-MS or LC-MS/MS capability (the five drugs identified in Table 5.6 for adult and child samples, all drugs and metabolites for neonatal samples, and for other samples for which there was insufficient hair to carry out a GC-MS test) was another significant step forward. ELISA was not designed to quantify drug or metabolite concentrations, and it should never have been used for that purpose.102

Table 5.6 Drugs and metabolites routinely tested by MDTL for adult and child samples, September 2010–April 2015

September 1, 2010–May 9, 2014

May 10, 2014–April 17, 2015

GC-MS

ELISA

(qualitative)

LC-MS/MS

ELISA

(qualitative)

Cocaine

Benzoylecgonine

Norcocaine

Cocaethylene

6-AM

Morphine

Codeine

Hydrocodone

Hydromorphone

Oxymorphone

Oxycodone

Methadone

Meperidine

Amphetamine

Methamphetamine

MDMA

MDA

Cannabis/THC

Benzodiazepines

Barbiturates

PCP

LSD

Cocaine

Benzoylecgonine

Norcocaine

Cocaethylene

6-AM

Morphine

Codeine

Hydrocodone

Hydromorphone

Oxymorphone

Oxycodone

Methadone

Meperidine

Amphetamine

Methamphetamine

MDMA

MDA

Cannabis/THC

Benzodiazepines

Barbiturates

PCP

LSD

Table 5.7 Drugs and metabolites routinely tested by MDTL for neonatal samples, September 2010–April 2015

September 1, 2010–May 9, 2014

May 10, 2014–April 17, 2015

ELISA (qualitative)

LC-MS/MS

ELISA

(qualitative)

Cocaine

Benzoylecgonine

Opiates

Amphetamines

Methamphetamines

Oxycodone

Methadone

Meperidine

Cannabis/THC

Benzodiazepines

Barbiturates

Phencyclidine

LSD

Cocaine

Benzoylecgonine

Norcocaine

Cocaethylene

6-AM

Morphine

Codeine

Hydrocodone

Hydromorphone

Oxymorphone

Oxycodone

Methadone

Meperidine

Amphetamine

Methamphetamine

MDMA

MDA

Cannabis/THC

Benzodiazepines

Barbiturates

Phencyclidine

LSD

99. Third, in September 2010, MDTL improved its reporting procedures by identifying in its results reports when a sample was referred to a reference laboratory for confirmation testing as well as the result obtained from that laboratory. This change improved transparency for the Laboratory’s customers and other users of its hair tests and ensured that there was no misunderstanding about which laboratory had carried out the test.

4.1.2 Written Standard Operating Procedures and Contemporaneous Documentation

100. In December 2010, MDTL adopted written standard operating procedures for its analytical methods as well as other policies and procedures on the non-analytical aspects of its operations (e.g., personnel, safety) as part of its quality management system.103 MDTL also implemented a corrective action / preventive action protocol to record any errors or anomalies in its analytical processes, ranging from batch fails to customer complaints, and to address and close out those issues.

101. The standard operating procedures set out in detail the steps to be followed in each case, from calibration and sample preparation through to quality control. In addition, MDTL began to keep better contemporaneous records of the steps taken in a particular case. For example, it began using a drug-testing log, which recorded who entered the data; who had cut, chopped (i.e., homogenized), and washed the hair; and when each of those steps was taken. The implementation of standard operating procedures and the drug-testing log represented a significant improvement in MDTL’s practices, removing a layer of uncertainty about the steps taken and the decisions made in a particular case.

102. Indeed, the Independent Review did not find in the post–September 2010 period many of the same inconsistencies it had previously found in MDTL’s practices:

  • The standard operating procedures set out the minimum sample volumes to be used for both the GC-MS and LC-MS/MS procedures.
  • MDTL’s standard operating procedures specified (for all adult samples, but not for neonatal and child samples) when samples would be washed before analysis and how.104 The drug-testing logs and results reports also indicated when a sample was washed before analysis.
  • The extraction methods that MDTL reported to the SoHT in its proficiency test submissions from 2011 to 2014 were consistent with the standard operating procedures that MDTL had in place at the time, suggesting that the Laboratory was using standardized extraction techniques before analysis.
  • MDTL’s standard operating procedures contained clear protocols for the use of quality controls: positive and negative control samples were run with every batch, and there were clear records showing when the controls passed or failed. In addition, MDTL implemented batch-to-batch monitoring by plotting the control results from each batch on a Levey-Jennings chart.
  • The standard operating procedures included defined cut-offs, which MDTL staff consistently applied.

103. That said, the Independent Review identified at least two aspects of MDTL’s analytical process, relating to the use of calibrators and the criteria for acceptability, that were not documented properly in its standard operating procedures. First, although the December 2010 standard operating procedure for MDTL’s ELISA–GC-MS method provided for a five-point calibration (i.e., the use of five calibrators for every batch), on December 1, 2011 (one year after the adoption of the procedure), MDTL reduced the calibration requirement to a one-point calibration at 4 ng/mg concentration, run twice. However, the Laboratory did not document the change in calibration methods until three years later – in November 2014, when the ELISA–GC-MS method was no longer in use (having been replaced by MDTL’s LC-MS/MS procedure).

104. Second, Paula Walasek, the laboratory technologist who was primarily responsible for carrying out the GC-MS tests, told the Independent Review that, when the acceptance criteria contained in the December 2010 ELISA–GC-MS standard operating procedure were not met initially, she used a technique called “background subtraction,” in which she manually removed interference from the background of the mass spectra to try to come up with a better match to meet the acceptance criteria. Although background subtraction is an acceptable technique in practice, that technique was not documented in MDTL’s standard operating procedure at all.

105. The fact that these alterations were made to the procedure in practice but not in writing raises concerns about what other steps may have been changed in the process without proper documentation. Laboratory processes must be validated and standardized both to ensure that the methods followed are appropriate and to avoid uncertainty or doubt about what steps were taken in a particular case. Accordingly, although the standard operating procedures that MDTL adopted in December 2010 and the Laboratory’s maintenance of contemporaneous records were a significant improvement on its previous practices, some aspects of the procedures remained deficient.

4.1.3 Improved Oversight

106. The provincial regulation and the Ontario Laboratory Accreditation requirements (which must be met in order for a laboratory to be licensed and accredited to offer clinical testing in the province) have defined roles for laboratory technicians and laboratory technologists. The former perform “laboratory tests which require limited technical skill and responsibilities” under direct supervision, whereas the latter perform “laboratory tests which require the exercise of independent judgment” under general supervision.105

107. Starting in September 2010, the GC-MS tests on drugs of abuse were primarily carried out by Ms. Walasek, who had the qualifications to be a laboratory technologist under the regulation. MDTL told the Independent Review that, once a GC-MS test was completed, the laboratory technologist who performed the test (usually Ms. Walasek) reviewed the test results and signed an internal document, called the “YES report,” which it downloaded by batch or by date. In addition, another individual – the quality manager (or her designate if she was not available), the laboratory manager, or another laboratory technologist – reviewed the quality control data for the batch before the result was reported to the customer, and every month the quality manager and MDTL’s laboratory supervisor reviewed MDTL’s quality control data to ensure that there were no anomalies in the results MDTL was generating.

108. Ms. Karaskov, who was not a laboratory technologist, continued to carry out the tests using ELISA. As with the GC-MS data, the quality control data for MDTL’s ELISA tests were reviewed by the quality manager (or her designate if she was not available), the laboratory manager, or a laboratory technologist before the result was reported. Although no laboratory technologist was involved in the analysis, a laboratory technologist signed the “YES report” for the ELISA batch.

109. This practice of having the laboratory technologist approve the result and having someone else review the quality control data for the batch before the result was reported was an improvement in MDTL’s previous practices. Among other things, it helped to ensure that the results reported to the customer were, in fact, the results the Laboratory had generated. In the later case files reviewed by the Independent Review, all the results that MDTL reported to the customers arising from its GC-MS analysis matched the results contained in the underlying data provided – unlike in the 2005–August 2010 cases described above in which there appeared to be transcription errors, switching of results, or even unknown results.106

110. Despite these improvements, however, MDTL still lacked the checks and balances required of a forensic laboratory. The Laboratory continued to rely on the person who carried out the test to review the final results before reporting the result. There was no routine review by anyone else of the underlying chromatograms or mass spectra for each GC-MS test that the Laboratory had run to ensure, for example, that the laboratory technologist who carried out the test had interpreted the data in accordance with MDTL’s acceptance criteria. The accepted practice in forensic laboratories is to have a second set of eyes on not only the quality control data but also on each peak in the underlying chromatogram and mass spectra, in order to confirm the accuracy and reliability of the analytical result before it is reported. Accordingly, although important improvements were made to MDTL’s oversight and reporting practices, even after September 2010 MDTL still did not maintain the level of supervision and oversight required of a forensic laboratory.

111. In addition, Dr. Koren continued to sign MDTL’s results reports until February 2013 (after which the reports were sent to customers unsigned), though he did not have any involvement in the underlying test or the review of the quality control data. Having the signature of the laboratory director on each results report issued until that date suggested that Dr. Koren had some involvement in, or at least had taken some steps to approve, the result. However, even after September 2010, that assumption was not true.

4.1.4 Routine Washing of Adult Samples

112. Under the December 2010 standard operating procedures, all adult samples were required to be washed twice with dichloromethane, an organic solvent, after segmentation and before testing. The SoHT has recommended washing samples with a combination of water (or an aqueous-based diluted buffer) and an organic solvent. Although MDTL did not use the former, its practice of two washes of dichloromethane was a dramatic improvement over not washing at all. Its washing protocol likely would have been sufficient to remove much of the external contaminants on the hair. MDTL’s implementation of a washing protocol for adult hair in the post–September 2010 period was one of the most significant improvements in its sample preparation process.

4.1.5 Complete Case Files

113. Many of the record-keeping deficiencies that the Independent Review identified in the 2005–August 2010 case files were not evident in the later cases that the Independent Review obtained. All the GC-MS and LC-MS/MS data were available electronically for all the cases requested. In addition, the later case files appeared to contain whatever contemporaneous records would have been available for the particular file. Although these improvements may have reflected the fact that the files were for samples that were tested more recently than the earlier files from 2005 to August 2010 (and were thus more readily retrievable), I consider the improved state of MDTL’s case files to be a reflection of the overall advancement in MDTL’s laboratory practices and quality management system from September 2010 onward.

4.2 Inadequacies in MDTL’s Analytical Methods

114. Despite all the above improvements, several aspects of MDTL’s analytical processes from September 2010 to April 2015 remained lacking and fell below internationally recognized forensic standards.

4.2.1 Deficiencies in GC-MS Procedure

115. The Independent Review identified several significant flaws in MDTL’s GC-MS procedure. In order to appreciate these flaws and their impact, it is important first to understand how the GC-MS instrument operates.

116. GC-MS works by using a gas to transport the sample through a long, thin column coated internally with a chemical material. Because the different compounds within a sample travel through the column at different rates, this process, called chromatography, separates the compounds contained in the sample from one another as they interact with the chemical coating inside the column. Once the separated compounds reach the end of the column, a mass spectrometer fragments the compounds into individual ions. Because molecules fragment in a predictable way under controlled conditions, the molecular properties of the fragmented ions create a “molecular fingerprint” that can then be used to identify the compounds contained in the sample.

117. The instrument creates two charts that are examined when identifying and quantifying a compound:

  • A chromatogram shows when the separated compounds reached the end of the column (referred to as the “retention time”). Each compound in the sample is represented as a peak on the chromatogram. Internal standards are used in every sample to establish the expected retention time for the drug as well as to correct for any potential differences in recovery arising from the sample preparation and extraction process. A peak on the chromatogram at the expected retention time would indicate the presence of a compound, and the area under the peak corresponds to the quantity of the compound within the sample.
  • A mass spectrum is created for each of the compounds that have been separated on the chromatogram. The mass spectrum plots the fragmented ions’ mass-to-charge ratio (referred to as m/z) and their intensity from low to high. An example of an electron impact mass spectrum for cocaine is found in Figure 5.1. As seen there, cocaine can be fragmented into many ions, the most significant ones being found at m/z 82, m/z 182, and m/z 303.107

118. The GC-MS instrument can be run in two different modes: “full scan” mode and “selected ion monitoring” mode. Under the full scan mode, the instrument compares the mass spectra from the sample tested with a “library” of mass spectra to try to find a match. The potential compounds that are identified by the software from the library search are assigned a percentage match (or a “similarity index” – the term used by the software for MDTL’s GC-MS instrument). This percentage indicates the extent to which the mass spectra in the sample resemble the reference spectra in the library. As a result of the manner in which the instrument operates, running the GC-MS instrument in full scan mode can be quite useful when identifying unknown compounds in a sample.

119. By contrast, selected ion monitoring searches for specific compounds by gathering data only for the ions that are of interest (referred to as diagnostic ions). Diagnostic ions are categorized into target (or quantifier) ions and qualifier ions: The laboratory measures the relative intensities of the qualifier ions to the target ion to confirm the presence of the target compound. Once identified, the target ion is then used to quantify the amount of the compound in the sample. When operating the GC-MS instrument in selected ion monitoring mode, the accepted practice is to identify a minimum of three diagnostic ions (one target ion and two qualifier ions) and assess their relative intensities before confirming the identity of the substance. Because the instrument is set to look for specific mass-to-charge ratio values, selected ion monitoring is more sensitive and can detect lower concentrations of compounds in a sample than is possible in an instrument set to full scan mode.

Figure 5.1 Example of an electron impact mass spectrum for cocaine

Figure 5.1 Example of an electron impact mass spectrum for cocaine

Source: Cocaine mass spectrum using Maurer, Pfleger, Weber Mass Spectral Library using electron impact GC-MS.

120. Turning to MDTL’s GC-MS procedure, the Laboratory’s December 2010 standard operating procedure included the use of deuterated internal standards for each drug and metabolite108 – the accepted practice. MDTL ran its GC-MS instrument in full scan mode. Nevertheless, MDTL listed the diagnostic ions it says it used for the 17 drugs and metabolites it tested for using its GC-MS procedure. They are set out in Table 5.8.

121. According to MDTL’s standard operating procedure, two checks were required before confirming the presence of the drug or metabolite:

  • The retention time of the peak on the chromatogram must not differ by more than 1 percent or +/– 0.2 minutes (whichever is smaller) from the retention time of the internal standard.
  • The diagnostic ions referred to in Table 5.8 must all be present on the mass spectrum, and the relative intensities of the diagnostic ions to one another must not differ by more than 35 percent relative to the reference spectrum from the library. If this requirement is met, the GC-MS instrument will return a match (or similarity index) of more than 65 percent.

Table 5.8 Diagnostic ions identified in standard operating procedure “MDTL 3500/01, Determination of Drugs of Abuse in Adult and Child Hair by ELISA-GCMS”

Drug or metabolite

Diagnostic ions (m/z)

Amphetamine

116

Methamphetamine

130

MDA

116

MDMA

130

Codeine

371, 146, 178

Hydrocodone

328, 297, 120

Oxycodone

416, 230, 401

Hydromorphone

386, 355

Oxymorphone

474, 287

6-AM

399, 340, 287

Morphine

429, 196, 236

Cocaine

182, 82, 303

Benzoylecgonine

240, 82, 361

Cocaethylene

196, 82, 317

Norcocaine

240, 346

Methadone

296

Meperidine

247

Note: Target (quantifier) ions are in bold; qualifier ions are not.

122. Ms. Walasek, who had primary responsibility for carrying out the GC-MS analysis for drugs of abuse, told the Independent Review that her practice was to visually check the retention time for the peak of interest on the chromatogram and then to examine the similarity index generated by the GC-MS software for the peak. Although the standard operating procedure provided for a similarity index of more than 65 percent, Ms. Walasek said that, in practice, she would look for a similarity index of more than 70 percent. If the similarity index generated by the instrument was more than 70 percent, she would consider that the peak had been confirmed for the particular drug or metabolite. If, however, the similarity index was less than 70 percent, she would manually review the mass spectrum for the peak to see if she could obtain a better match. Because other ions may exist in the “background” and may interfere with a peak on the mass spectrum, she would sometimes use the instrument to subtract the background ions to see if she could obtain a better match with the library. Often, after subtracting the background ions, she could obtain a 65 percent or greater similarity index match to confirm the identity of the drug or metabolite.

123. There are several deficiencies in the manner in which MDTL carried out its GC-MS analysis from September 2010 to May 2014, based on both MDTL’s standard operating procedure and the process Ms. Walasek described. These deficiencies relate to MDTL’s ability to identify the presence of a drug or metabolite reliably and its ability to quantify that drug or metabolite reliably.

4.2.1.1 Identification Method

124. MDTL’s GC-MS procedure did not meet the accepted practices for GC-MS confirmation in several respects. First, the accepted practice when relying on results obtained by the GC-MS instrument in full scan mode is to require a match of 95 percent or higher. MDTL’s procedure, which required a similarity index of only 65 percent or higher, was therefore inadequate. MDTL’s use of the lower similarity index created a risk that the appropriate identification criteria for the drug or metabolite (including the assessment of the ion ratios from the target and qualifier ions) were not met.

125. In fact, it is unclear why MDTL included a list of diagnostic ions in its GC-MS procedure (see Table 5.8) if it was running the GC-MS instrument in full scan mode. When the instrument is run in full scan mode, the percentage match (or similarity index) reflects the match between the compounds separated in the sample and those contained in the instrument’s library. The selection of diagnostic ions applies when the instrument is operated in selected ion monitoring mode (the diagnostic ions being those ions that the instrument are selected to monitor). Because MDTL relied on the automatically generated similarity index, the Independent Review was unable to verify how the match was made.

126. In any event, to the extent that MDTL was monitoring for the diagnostic ions set out in Table 5.8, the accepted practice before a positive identification can be made is that at least three diagnostic ions (one target ion and two qualifier ions) must be identified, and their relative intensities must match those contained in the library or the internal standards. However, as set out in Table 5.8, MDTL used only

  • one diagnostic ion (the target ion) to identify six of the drugs: amphetamine, methamphetamine, MDA, MDMA, methadone, and meperidine; and
  • two diagnostic ions (one target ion and one qualifier ion) to identify three of the drugs: hydromorphone, oxymorphone, and norcocaine.

127. To the extent that MDTL used only one diagnostic ion (the target ion) in its confirmation tests, its practice fell below the accepted standards for GC-MS confirmation.109 It created a risk that samples reported as positive for those drugs were not sufficiently identified in those samples; in other words, a false positive.

128. Second, even putting aside the sufficiency of the similarity index, it was inappropriate for MDTL to rely solely on the GC-MS instrument to identify the drug or metabolite. When identifying compounds using mass spectrometry, the analyst must not only identify the diagnostic ions and examine their relative intensities but also ensure that there is an absence of other significant ions on the mass spectra that might suggest the presence of something else.110 The presence of another significant ion would alert the analyst to the fact that the substance may not be the target drug or metabolite. By relying solely on the similarity index for the diagnostic ions generated by the GC-MS instrument, MDTL did not take this important step to ensure the accuracy of its identification.

129. Third, while using background subtraction to remove any significant ions near the peak of interest is an accepted practice when the requisite match criteria have not been satisfied, this approach was not contained in MDTL’s standard operating procedure, and there were no guidelines to establish when background subtraction should be used or how. Nor was the ultimate match obtained as a result of the background subtraction recorded. As a result, in the many cases obtained where the similarity index generated by the instrument was less than 65 percent, but a drug or metabolite was positively identified, the Independent Review was unable to verify that the drug or metabolite reported was indeed confirmed by this background subtraction approach.

130. MDTL’s failure to record the background subtraction that ultimately resulted in the match is further compounded by the fact that there was no checking of the mass spectra to ensure that MDTL’s acceptance criteria had been met before the result was reported. As a result, the only person who could confirm that a sufficient match had been obtained after using background subtraction was the laboratory technologist who carried out the test.

4.2.1.2 Quantification Method

131. On December 1, 2011, MDTL moved from the five-point calibration that was documented in its standard operating procedures to a one-point calibration at 4 ng/mg, which was run in duplicate, to improve its turnaround times. This change meant that MDTL tested the same calibrator (at a concentration of 4 ng/mg) twice within the same batch.111

132. However, a review of MDTL’s validation data used to justify the change revealed that the calibration curve that MDTL generated was not sufficiently linear to justify the Laboratory’s use of a five-point calibration, let alone a one-point one. Without a sufficiently linear calibration curve, the concentrations that MDTL generated using its GC-MS tests were inadequate and unreliable for forensic purposes.112

4.2.2 Continued Reliance on Unconfirmed ELISA Results

133. Although MDTL was right to stop reporting ELISA results quantitatively in September 2010, it also should have refrained from reporting its results qualitatively. What MDTL should have done was refer every preliminary positive result to its reference laboratory for confirmation before reporting any result to the customer. Alternatively, MDTL should have made it clear in its results reports that the result obtained from its ELISA test was only preliminary in nature. Such a caveat would have alerted users of its results reports to the limitations in its ELISA-only results and given them the opportunity to request a confirmation test if desired. However, the Laboratory took neither of these steps.

134. In fact, the manner in which MDTL reported its results was confusing and likely would have led users of the test results to believe that the ELISA result was more reliable than it really was:

  • Although MDTL used the term “screen” in its results reports when referring to ELISA results that were also confirmed by GC-MS, it did not use the term when reporting its unconfirmed ELISA results for the drugs it did not test using GC-MS. For example, in one case MDTL stated that it had carried out a screen for amphetamines and opiates, but in the same report, it stated its ELISA results for cannabis and benzodiazepines without indicating that those were also screens. Reporting the results in this manner indicated that the ELISA tests for cannabis and benzodiazepines were not from screens but that the ELISA tests for amphetamines and opiates were from screens, suggesting that the former results were more reliable than the latter.
  • The Independent Review found some cases in which MDTL reported the ELISA result for benzodiazepines and cannabis as “> 0.20 ng/mg.” MDTL used a 0.2 ng/mg cut-off for benzodiazepines and cannabis. Although it may be true that a positive result was one that was greater than 0.2 ng/mg, reporting the results in this manner suggested to the user of the results report that a quantitative result had been obtained using the ELISA test, though it had not.
  • MDTL’s use of the term “trace” for its ELISA-only results was also inappropriate. This term is most commonly associated with a confirmed test result that is at a concentration which cannot be accurately measured. MDTL’s use of the term for its ELISA-only results was misleading because it suggested that the drug or metabolite was present, though it had not, in fact, been confirmed.

135. Accordingly, MDTL’s practices in continuing to rely on ELISA results – albeit qualitatively, rather than quantitatively – continued to fall short of internationally recognized forensic standards.

4.2.3 Routine Failure to Wash Neonatal and Child Samples

136. Even in the post–September 2010 period, MDTL did not wash child (under 16 years of age) and neonatal (under 6 months) samples before analysis. MDTL drew this distinction because, for child and neonatal samples, it considered that it was important to detect exposure (not necessarily use) of a drug or metabolite, and because (given the smaller sample size) the instrument might not have been sensitive enough to detect the small amounts of drug or metabolite that might be in the hair.

137. If MDTL was testing hair solely to detect exposure to (rather than use of) a drug, in theory its decision against washing the hair would not have been problematic as long as it made it clear to customers and other users of its test results that the sample was not washed, as well as its reasons for deciding not to wash the sample and the implication of this decision. In practice, however, MDTL’s methods were inappropriate given the interpretations that the Laboratory provided: It suggested that conclusions could be drawn about the frequency of a caregiver’s use or the level of risk in the child’s environment based on a child’s, or neonate’s results.

138. Moreover, MDTL’s results reports did not make it clear to the user of the test results that the samples were not washed. In contrast to the results reports that indicated that the adult samples were “washed” before testing, there was no similar notation on the reports to indicate that the child and neonatal samples were “unwashed.” Without this indication, MDTL failed to alert the users of the test results that the samples were not washed, along with the implications of its decision not to do so.

4.2.4 Inadequate Chain-of-Custody Procedures

139. The concerns about chain of custody that I identified for the 2005–August 2010 period continued to apply in the September 2010–April 2015 period. For example, although a record was kept of the cutting, chopping, and washing steps, transfers within the Laboratory were not witnessed and documented. It is clear that chain of custody was still not a concern or a priority for MDTL once the sample was received by the Laboratory.

140. In 2013, MDTL moved from SickKids’s main hospital building to the Peter Gilgan Centre for Research and Learning. The Independent Review visited the Laboratory on several occasions. Although the facilities appeared to be state of the art, they lacked the security features of a forensic laboratory. Researchers and students were permitted unfettered access to the laboratory benches where MDTL received the samples, prepared them for analysis, and carried out the analysis. Visitors were required to sign into the building and needed a pass to access the Laboratory, but no records were kept of unauthorized people as they came into and left the Laboratory. Although MDTL’s procedures were much improved in the period from September 2010 to April 2015, MDTL remained a clinical laboratory within a hospital setting.

5. Alcohol Markers: 2005–15

141. MDTL tested for fatty acid ethyl esters (FAEEs) in-house starting in March 2004.113 After February 2011, when a sample tested above the FAEEs cut-off, MDTL began routinely to refer samples to its reference laboratory for ethyl glucuronide (EtG) testing. It continued to do so until the Laboratory closed.

142. MDTL did not provide the Independent Review with sufficient materials to assess the adequacy and reliability of the FAEEs tests it carried out before June 2007. However, the Independent Review’s assessment of the post–June 2007 documents and case files revealed that MDTL’s FAEEs analysis during that later period did not suffer from many of the deficiencies identified in its analytical methods for drugs of abuse from 2005 to August 2010. That said, several shortcomings remained in MDTL’s hair analysis for alcohol markers during the period under review.

143. In addition, because alcohol marker testing is a relatively new field in hair testing, and the science of hair testing for alcohol markers has evolved over the past 10 years, my findings about the adequacy and reliability of MDTL’s hair tests for alcohol markers are made with the benefit of hindsight. Some aspects of MDTL’s practices that complied with the standards at the time (the use of non-standard hair lengths and reliance on FAEEs analysis alone) would not be considered reliable and adequate for forensic purposes according to today’s standards. These issues are described in greater detail below.

5.1 Evolution in the Science of Alcohol Marker Testing in Hair

144. Alcohol marker testing is a relatively recent addition to the field of hair testing, having gained prominence only in the last decade or so. The SoHT released its first consensus statement with respect to alcohol marker testing in hair in 2009. The science of alcohol marker testing in hair has evolved in three central ways from the mid-2000s to today.

145. First, FAEEs were the initial alcohol markers to be recognized for hair. However, over the last decade, studies began to show that ethanol-containing hair products can create FAEEs in hair by reacting with fatty acids in the sebum, sometimes dramatically increasing the concentration of FAEEs detected in the hair test. As a result, individuals who use ethanol-containing hair products may, even when they abstain from alcohol, test positive for FAEEs. In the mid- to late 2000s, it became standard practice for hair-testing laboratories to request a list of hair products (and their ingredients) used during the period under analysis so they could be taken into account with a test result.

146. Second, since 2009, the SoHT has developed cut-offs to assess chronic excessive alcohol consumption and, more recently, abstinence.114 The use of internationally recognized cut-offs to assess the extent of a person’s alcohol use is unique to alcohol marker testing, and there are no similar cut-offs for drugs of abuse. However, only two set lengths are recommended to be used for FAEEs testing: a 3 cm or a 6 cm length, measured from the scalp (referred to as 0–3 cm and 0–6 cm, respectively).115 In recent years, the SoHT has developed different cut-offs for each of these lengths:

  • In its 2009 consensus statement, the SoHT recommended using a cut-off of 0.5 ng/mg for the 0–3 cm length for FAEEs to assess chronic excessive alcohol consumption.
  • In 2011, the SoHT released an updated consensus statement that contained different cut-offs for FAEEs, depending on the length of the hair used.116 The SoHT cut-off to assess chronic excessive alcohol consumption for a 0–3 cm length was 0.5 ng/mg, and the cut-off for a 0–6 cm length was 1.0 ng/mg. Because EtG does not incorporate into hair in the same way, the SoHT recommended the same cut-off for EtG for both the 0–3 cm and the 0–6 cm segments (30 pg/mg). The SoHT also made it clear that any samples under 3 cm in length should be interpreted with caution.
  • At its June 2014 meeting, the SoHT again updated its consensus statement, this time to include cut-offs for abstinence assessment for both FAEEs and EtG.117 The abstinence cut-offs were 0.2 ng/mg for a 0–3 cm segment and 0.4 ng/mg for a 0–6 cm segment for FAEEs (although the SoHT also made it clear that the analysis of FAEEs alone was not recommended to determine abstinence), and 7 pg/mg for EtG. Accordingly, since 2014 the SoHT has provided cut-offs to assess both abstinence and chronic excessive alcohol consumption, but nothing in between.118

147. Third, around the mid-2000s, EtG became known as another alcohol marker in hair and, in recent years, has become widely accepted as the more reliable alcohol marker in hair. In fact, since the early 2010s, given the effect of hair product interference, many hair-testing experts have considered FAEEs to be an unreliable method to assess chronic excessive alcohol consumption and have stopped testing for FAEEs entirely, relying solely on EtG (which is unaffected by ethanol-containing hair care products).

5.2 Evolution in MDTL’s Analytical Methods

148. Although MDTL started testing hair for drugs of abuse in child protection and criminal cases as early as 1999, it did not begin testing for FAEEs in hair until several years later, in March 2004. However, the methods and instruments that MDTL used to test for FAEEs changed over time.

149. The Laboratory offered alcohol marker testing to customers from March 2004 to November 2004.119 However, MDTL was unable to identify which instrument it used and what method it followed during this time. Nor was the Laboratory able to provide the Independent Review with any case files from the tests it carried out during this period. Because no ELISA kits are commercially available for FAEEs, the analysis would have been performed on a GC-MS instrument.

150. In any event, MDTL advised the Independent Review that, from December 2004 to December 2005, it sent hair samples to another laboratory within SickKids for analysis.120 Again, however, MDTL was unable to verify what procedure was followed by this other laboratory. Nor was MDTL able to ascertain if all hair samples were sent to the other laboratory at SickKids for FAEEs analysis during this time or if MDTL continued to carry out any FAEEs analysis in-house.121

151. In October 2005, MDTL sent a total of 15 hair samples that had been tested for research purposes (seven samples) and what it called “clinical” purposes (eight samples) to Professor Fritz Pragst’s laboratory at the Institute of Legal Medicine and Forensic Sciences, University Hospital Charité, in Berlin for comparison.122 Although MDTL could not identify whether those 15 samples had been tested in-house at MDTL or if they had been tested at the other laboratory at SickKids, the results of the comparison analysis were troubling. According to MDTL, the FAEEs test that had been carried out either at MDTL or the other laboratory at SickKids yielded a positive result (i.e., above the SoHT cut-off of 0.5 ng/mg) in all eight of the “clinical samples.” However, only two of those samples tested positive at Professor Pragst’s laboratory. Accepting Professor Pragst’s results, MDTL may have reported false positives to its customers in three-quarters of those cases. Mr. Gareri could not recall whether corrections or notices pertaining to these samples were issued to customers at the time. Ultimately, MDTL decided to stop routine FAEEs hair analysis until what MDTL called “methodological issues” were resolved.

152. Accordingly, from January 2006 to May 2007, no FAEEs hair tests were carried out by MDTL or any other laboratory within SickKids. MDTL told the Independent Review:

  • During this time, MDTL advised customers in the presentations it gave to child protection agencies that FAEEs analysis was not available. In addition, in August 2007 MDTL’s customers were advised that, for 43 samples collected from February to December 2006, FAEEs analysis was unavailable, so no result was reported.
  • Despite telling customers that FAEEs analysis was unavailable, MDTL referred165 samples to Professor Pragst’s laboratory for FAEEs analysis from January 2006 to March 2007, and MDTL reported the results of that laboratory.123 MDTL told the Independent Review that the referral of these samples to Professor Pragst’s laboratory seemed to be conducted on a “special case-by-case basis,” and Mr. Gareri could not recall how the selection was made. MDTL did not identify on its results reports that the sample had been tested by another laboratory.

153. In the meantime, in November 2006 MDTL sent its research associate, Dr. Katarina Aleksa, to Professor Pragst’s laboratory to learn the FAEEs method. When Dr. Aleksa returned from Berlin, she adopted the method she had learned from Professor Pragst, and a new GC-MS instrument was eventually purchased by MDTL and used for FAEEs analysis.124 MDTL began to offer hair testing for alcohol markers under this new methodology in June 2007. MDTL continued to use the Pragst method for hair analysis for FAEEs until it shut down its operations in April 2015.

154. In February 2011, because it did not have the capability to do so in-house, MDTL also began to send all the samples that were above the SoHT cut-off for FAEEs (and for which there was sufficient sample) to its reference laboratory, USDTL, for EtG analysis. Consistent with its post-2010 practice for drugs of abuse, MDTL indicated on its results reports that the sample had been referred to USDTL and reported both its FAEEs result and the USDTL result for EtG.

5.3 Adequacy and Reliability of MDTL’s Analytical Methods

155. MDTL was unable to provide the Independent Review with any case files or other documents for the FAEEs method that either it or the other laboratory at SickKids used from March 2004 to December 2005. Accordingly, the Independent Review was unable to assess the adequacy and reliability of the methods used during the part of this period that is within the Independent Review’s mandate.125 By failing to maintain documents that outlined the analytical steps in its FAEEs analysis, MDTL fell below internationally recognized forensic standards that required forensic laboratories to have standard operating procedures and record retention policies.

156. However, the Independent Review was able to assess the FAEEs tests that MDTL carried out from June 2007 to April 2015 under the Pragst method. With some exceptions, described below, the Independent Review found MDTL’s analytical procedure for FAEEs testing under the Pragst method to be a fairly reliable method.

5.3.1 Improvements over Analytical Methods for Drugs of Abuse

157. Several of the deficiencies that plagued MDTL’s drug tests in the 2005–10 period did not affect its FAEEs hair tests. First, MDTL had a standard operating procedure for FAEEs testing as early as April 2008 – in contrast to drugs of abuse, where it did not implement standard operating procedures until December 2010.126 Moreover, although MDTL did not have a standard operating procedure from June 2007 until April 2008, the procedures for FAEEs analysis were documented, though not formalized. Likely because of the existence of such procedures, the Independent Review did not find the same inconsistencies in sample preparation and use of quality controls for FAEEs that it identified in the 2005–August 2010 period for drugs of abuse.

158. Second, in contrast to the samples that were tested for drugs of abuse from 2005 to August 2010, MDTL routinely washed all the samples it tested for FAEEs under the Pragst method. Moreover, at least as early as July 2007, it recorded in its results reports that the sample was washed. There is wide acceptance that samples must be washed before FAEEs testing, and MDTL complied with that standard.

159. Moreover, it is important to recognize that, unlike its hair tests for drugs of abuse, in some respects MDTL’s methods for alcohol markers stayed current with the evolving science. Given the evolving science, MDTL’s decision in February 2011 to refer all future samples that fell above MDTL’s cut-off for chronic alcohol abuse to USDTL for EtG testing was appropriate.127 There is no reason to believe that any EtG test conducted by USDTL testing methodology was not reliable for forensic purposes.

5.3.2 Deficiencies in Analytical Methods for FAEEs

160. Despite the clear improvements MDTL practised in its FAEEs analysis, compared with its testing of drugs of abuse, the FAEEs method still fell below the accepted practice in several ways. First, the FAEEs method exhibited similar shortcomings to those seen in the GC-MS method for drugs of abuse:

  • For FAEEs, MDTL ran the GC-MS instrument in full scan mode, requiring only a 60 percent similarity index. Again, for the reasons set out above, that match percentage was insufficient.
  • Although MDTL’s more recent standard operating procedures for FAEEs analysis provided for background subtraction, they did not give sufficient guidance to the analyst on the acceptance criteria to be followed when subtracting background ions to search for a match. That absence led to the same issues with uncertainty of results that were identified above for drugs of abuse.
  • As with its GC-MS procedure for drugs of abuse, starting in 2008 MDTL used a one-point calibration for its FAEEs analysis. A review of the validation data for MDTL’s use of a one-point calibration revealed the same lack of linearity as was found in the validation data for the GC-MS procedure for drugs of abuse, thereby calling into question the quantitative reliability of the FAEEs results that MDTL generated. (However, in contrast to its GC-MS procedure for drugs of abuse, MDTL appropriately documented the one-point calibration in its standard operating procedure.)

161. I note that these shortcomings in MDTL’s procedure were not specified in the Pragst method, which did not contain steps to this level of detail. The Pragst method is widely accepted as an appropriate method for FAEEs analysis but leaves details of this nature to the individual laboratory.

162. Second, it seemed that a compound interfered with one of the FAEEs – ethyl stearate – in the chromatograms that the Independent Review examined. As a result of this interference, the peak for ethyl stearate on the chromatogram was not completely separated from the peak caused by the interfering compound. This interference can lead to issues with both identification (the compound is identified by the retention time for the peak) and accurate quantification (the compound is quantified by calculating the area under the peak). That said, ethyl stearate is the FAEE typically found in the lowest concentration of all the FAEEs that are tested in hair. Because the sum of four FAEEs is used to calculate the final concentration, the failure to adequately separate the ethyl stearate in the chromatogram likely would not have had a material impact on the results reported by MDTL.

163. Third, in one of the batches that was reviewed, the peaks on the chromatogram for the internal standards that were run with the samples tested varied significantly – in that case, up to 10-fold.128 Although it is not uncommon for internal standard peak areas to vary on account of slight changes in laboratory conditions, a 10-fold difference across the internal standards run in the same batch is alarming. It suggests that there were significant changes in sample preparation and/or that the efficiency of the extraction methods varied across the samples, both of which raise questions about the robustness and validity of MDTL’s methods.

164. Fourth, MDTL misapplied the cut-offs in two material respects:

  • Although the SoHT has recommended FAEEs cut-offs to assess chronic excessive alcohol consumption since 2009 and abstinence since 2014, MDTL used cut-offs that were not justified by the SoHT consensus statements or the academic literature: it created its own cut-offs to describe what it called “social drinkers” or “moderate / non-drinkers.” However, it was not appropriate to create categories in this manner.
  • Although the SoHT recommended a cut-off of 30 pg/mg for EtG to assess chronic excessive alcohol consumption, MDTL applied a 20 pg/mg cut-off. Again, the use of cut-off in this manner was not supported by either the SoHT consensus statements or the academic literature at the time. For results between 20 and 30 pg/mg, for example, MDTL would have interpreted the result as evidence of chronic excessive alcohol use, though the consensus (as reflected in the SoHT statement) would not have interpreted the result in this way.

165. Accordingly, MDTL’s methods for FAEEs were an improvement over those it used for drugs of abuse. However, the remaining deficiencies call into question the adequacy and reliability of the results. In addition, although USDTL conducted the EtG testing for MDTL, the Laboratory misapplied the SoHT cut-offs for those results.

5.3.3 Changes in the Science

166. In addition, there are two ways in which MDTL’s hair tests are inadequate and unreliable by current standards, but perhaps not by the standards of the day. First, the Independent Review identified samples in which MDTL used non-standard hair lengths to test for FAEEs. For example, in a case from 2005, MDTL interpreted the FAEEs results from three different samples obtained from the same person over a period of 1.5 years. The sample lengths varied from 9 cm, to 6 cm, to 4 cm in length (the last sample was segmented into two segments of 2 cm each). As the sample length can affect the amount of FAEEs that incorporate into hair because of the sebum that remains on the hair, the use of standard sample lengths (of 3 cm or 6 cm) is now the accepted practice. However, it was not until the late 2000s that the standardization of sample lengths became more widely recognized, and not until 2011 that the SoHT updated its consensus statement to include different cut-offs for the 0–3 cm and 0–6 cm lengths. Accordingly, although MDTL’s practices in using inconsistent sample lengths during the period covered by the Independent Review were not necessarily deficient when considered in the context of the accepted practice at the time, they may have affected the results that the Laboratory reported, and they fall below the accepted practice when viewed by today’s standards.

167. Second, many hair-testing experts have now called into question the usefulness of FAEEs testing, given the potential impact of ethanol-containing hair products on the results. Many experts now believe that FAEEs testing, on its own, is unreliable, and that EtG is, by far, the more reliable alcohol marker to test for in hair. Accordingly, although MDTL met the standards of the time by testing for FAEEs in hair, I am not satisfied that, in the many cases where only FAEEs results were obtained, the results would be adequate and reliable for forensic purposes. The current best practice is to test hair using EtG, and to assess the results according to the SoHT recommended cut-offs.


Notes

66 One of the experts retained by the Independent Review, Dr. Gail Audrey Ann Cooper, contributed to the creation of the 2012 consensus statement for drugs of abuse and to the 2011 and 2014 consensus statements for alcohol markers. Dr. Cooper has been a member of the board of directors of the SoHT since 2010.

67 The United Nations International Drug Control Programme is mandated to assist member countries in the development of national programs to combat drug trafficking and drug abuse and to educate national laboratories on recent developments in the science.

68 Professor Olaf H. Drummer, another expert retained by the Independent Review, contributed to both the 2001 and 2014 UN Guidelines.

69 Dr. Cooper also contributed to the Forensic Toxicology Laboratory Guidelines published by the United Kingdom and Ireland Association of Forensic Toxicologists in 2010. In addition, various organizations have created forensic toxicology guidelines specific for workplace drug testing (e.g., the European Workplace Drug Testing Society) and doping control in sports (e.g., the World Anti-Doping Agency).

70 The requirements to obtain forensic laboratory accreditation worldwide are generally based on the International Standards Organization (ISO), ISO/IEC 17025:2005, General requirements for the competence of testing and calibration laboratories (last reviewed in 2010).

71 The ELISA test for opiates detected a range of opiates, including morphine, codeine, and 6-AM (the primary metabolite of heroin). The ELISA test for methamphetamine detected both methamphetamine and MDMA. The ELISA test for amphetamine detected both amphetamine and MDA.

72 For example, the 2001 UN Guidelines made it clear that “results from immunoassays must be confirmed by a chromatographic method or any other independent technique providing equivalent selectivity.” Likewise, the SoHT’s 2004 consensus statement for the examination of hair for drugs provided: “Analytes of interest must be identified to minimize false negatives.”

73 It is an accepted and indeed common practice for a laboratory to send samples to another laboratory (the reference laboratory) for testing. Such referral typically occurs when a laboratory does not have the equipment or the expertise to carry out the test that has been requested. It may also occur as a way of double-checking the laboratory’s own results.

74 Although some ELISA kits are better than others at targeting the specific drug or metabolite (e.g., the ELISA tests for cocaine and benzoylecgonine have a lower rate of cross-reactivity than the test for opiates because of their relatively unique structures), the international standards are clear that, even for such drugs or metabolites, confirmation testing is required, not optional.

75 Neither of the experts retained by the Independent Review had ever heard of ELISA being used for quantitative purposes by any forensic toxicology laboratory around the world.

76 Before ELISA, MDTL used another immunoassay-based test known as radioimmunoassay.

77 Ms. Klein also noted that the ELISA and GC-MS results could not be expected to be exactly the same given the different methods of quantitation and extraction used.

78 MDTL told the Independent Review that it had found test results from Psychemedics, but it did not have the corresponding ELISA test results or comparison information for the samples that were sent to Psychemedics.

79 As set out in Chapter 2, concerns raised by the Broomfield case in the fall of 2014 led to my appointment as Independent Reviewer.

80 As described below, I have concerns with the reliability of MDTL’s GC-MS procedure. Accordingly, in some respects the comparison that MDTL conducted for the 164 samples involves a comparison of two unreliable methods.

81 False positives were defined as those samples that had tested positive using ELISA, but negative using GC-MS or LC-MS/MS. False negatives were defined as those samples that had tested negative using ELISA, but positive using GC-MS or LC-MS/MS. The Independent Review used MDTL’s cut-offs in assessing the rate of false positives and false negatives.

82 Because USDTL did not report concentrations above 10 ng/mg for cocaine and above 1.00 ng/mg for benzoylecgonine (and a large proportion of the samples that were tested had results above these thresholds), the Independent Review did not compare the ELISA versus LC-MS/MS quantitative results for the 53-sample data set.

83 The development of the “Motherisk Drug Testing Laboratory Procedures” was prompted by the circumstances surrounding Ms. Klein’s sudden departure in 2005, as well as the departure of the senior secretary in SickKids’s Research Institute, both of which involved allegations of fraud and gave rise to concerns about financial aspects of MDTL’s operations.

84 Testing long samples in bulk can have a dilution effect, as the larger volume dilutes the amount of drug that can be detected in the sample – potentially causing a false negative or falsely low result.

85 As described below, MDTL routinely washed hair samples for FAEEs analysis. The one exception was from 2010, where a sample that was collected and tested in August 2010 (shortly before MDTL moved to its new testing method) was described as being “washed” for all tests completed, and not only for FAEEs analysis.

86 Although laboratories carrying out tests for strictly research purposes may choose to adjust their methods in a specific test to optimize extraction, standardization is required in both forensic and clinical settings.

87 Because ELISA responses will change as a result of minor alterations in how the test is run from batch to batch (e.g., temperature fluctuations or differences in incubation times), it is entirely inappropriate to use the calibration curve from one ELISA batch to calculate the results for samples tested in another batch.

88 Although this dilution technique is an accepted practice when using confirmation techniques such as GC-MS or LC-MS/MS, it is inappropriate for immunoassay-based tests such as ELISA because dilution of the sample can materially change the manner in which the ELISA test responds to the sample.

89 According to Ms. Klein, before she left MDTL in April 2005, she checked Ms. Karaskov’s work.

90 The higher incidence with cocaine and benzoylecgonine likely reflects the greater number of cocaine and benzoylecgonine positive samples. As described above, cocaine and benzoylecgonine were among the most frequently tested drugs and metabolites from 2005 to August 2010.

91 MDTL did not disclose in its reports or evidence to the courts that three tests had been completed. Significantly, because MDTL did not locate them at the time, the six spreadsheets (three for cocaine and three for benzoylecgonine) were not made available to the Crown and defence counsel on the Broomfield appeal.

92 Similarly, the 2001 UN Guidelines require that “[a]ll hair samples undergo a decontamination procedure.” Although there is no standard washing protocol, the accepted approach at the time was to wash the hair with both an organic solvent and a water-based solution.

93 An analysis of the washings is considered to be a helpful tool in distinguishing between active drug use and passive exposure to drugs.

94 As described in Chapter 7, for the most part MDTL’s proficiency test submissions from 2001 to 2009 did not reflect its own analytical testing methods or results.

95 MDTL said it did not consider it necessary to have chain-of-custody documentation for samples collected within SickKids because the Hospital maintained custody of the sample.

96 Before this request, the Toronto Police Service had obtained a production order in May 2012 to compel SickKids to pass over “all protocols pertaining to the analysis of hair for evidence of cocaine and/or cocaine metabolites.”

97 The Independent Review learned that the accused’s sample in Broomfield was sent to USDTL for confirmation for 6-AM and was returned negative. However, the result was not reported to the police or the Crown (indeed, the results report that contained the GC-MS confirmed (negative) result specified that it was “for lab only”). MDTL could not explain why it decided to refer out the accused’s sample for 6-AM for confirmation, but not the child’s sample for cocaine or benzoylecgonine.

98 MDTL continued to use ELISA kits as a preliminary screen for cocaine, benzoylecgonine, opiates, oxycodone, methadone, meperidine, amphetamine, and methamphetamine.

99 The Independent Review did not identify any cases after May 2014 in which MDTL reported the ELISA result as “trace.”

100 As described below, however, the Independent Review identified several deficiencies in how MDTL carried out its GC-MS analysis from September 2010 to May 2014. Though significant, those deficiencies pale in comparison to MDTL’s previous ill-advised practice of reporting unconfirmed ELISA results for quantitative purposes.

101 Although certain elements in MDTL’s LC-MS/MS method could have been improved, the Independent Review did not identify the same number of deficiencies in the LC-MS/MS method that it found in MDTL’s GC-MS procedure.

102 However, as described below, MDTL’s practice of reporting the qualitative ELISA result still did not conform to internationally recognized forensic standards.

103 MDTL adopted some of its non-analytical policies and procedures in the months leading up to December 2010.

104 However, the standard operating procedures did not specify that the washings should be stored for later analysis if required (the SoHT’s recommended practice).

105 Laboratories, RRO 1990, Reg 682, s 1.

106 However, there was one case that the Independent Review identified in the post–September 2010 period where an ELISA result for cannabis / THC was reported as “> 0.20 ng/mg” but was actually negative.

107 The ion at m/z 303 represents the entire structure of cocaine, which occurs when the cocaine molecule remains intact despite the fragmentation process.

108 Deuterated internal standards refer to the use of standards in which the hydrogen in the drug or metabolite compound has been replaced with deuterium.

109 The accepted practice is not only to identify the three diagnostic ions but also to assess their relative intensities to one another. Where MDTL used fewer than three diagnostic ions, it could not have assessed those relative intensities to identify the compound.

110 Other significant ions may come from a co-eluting substance, for example, which may artificially increase the calculated concentration if the drug is actually present.

111 Running a calibrator in duplicate would allow MDTL to check that it was getting a consistent response for the same calibrator.

112 I note that MDTL’s quality control methods during this time serve to mitigate in some respects the risk of an inaccurate quantification for values in and around the control concentrations. However, because only two controls were run with each batch, any changes to the linearity of the calibration curve at the extreme ends (i.e., low and high concentrations) would not have been picked up by the controls.

113 However, there was a period in which MDTL referred hair samples to another laboratory for FAEEs analysis, as well as a period in which it suspended its FAEEs analysis altogether.

114 Chronic excessive alcohol consumption was defined by the World Health Organization as drinking in excess of 60 grams of pure ethanol (or approximately four to five drinks) per day.

115 Because FAEEs incorporate into hair primarily through sebum, and FAEEs from sebum that stays on the hair continue to diffuse into the hair even after it has grown out of the scalp, FAEEs concentrations tend to increase from the root end to the distal end of the hair.

116 The SoHT’s 2011 consensus statement followed a decision of the High Court of Justice Family Division in the United Kingdom that criticized laboratories involved in the analysis and interpretation of alcohol markers for their use of non-standard hair lengths, among other things: London Borough of Richmond v BWBCB & CB (through their Children’s Guardian), [2010] EWHC 2903 (Fam).

117 A result that is less than the cut-off would indicate abstinence.

118 Although the SoHT continues to recommend both FAEEs and EtG as independent markers for chronic excessive alcohol consumption, in its 2014 consensus statement it made it clear that FAEEs should not be used on their own to determine abstinence from alcohol.

119 MDTL told the Independent Review that the first FAEEs hair test result that was reported in its database was from December 2003. The second one was reported in March 2004, and analysis of hair samples for FAEEs appeared to be reported routinely following this date.

120 Mr. Gareri recalled that MDTL encountered some issues with the Laboratory’s GC-MS instrument – and those problems may explain the switch to the other laboratory.

121 The results reports that MDTL issued during this time did not indicate that the FAEEs test had been performed by another laboratory.

122 Professor Pragst is widely recognized as a leader in the field of alcohol marker testing in hair.

123 The 165 samples were collected between November 2005 and May 2007.

124 I note that, although MDTL acquired a GC-MS instrument in the laboratory for FAEEs analysis, it could not be used to test for drugs of abuse without proper validation.

125 Although MDTL provided the Independent Review with a copy of a graduate student’s thesis from 2004 that described FAEEs analysis, it was unable to confirm if the method described in the thesis was the method that MDTL used for its FAEEs analysis at this time. In addition, MDTL was able to locate two documents on Ms. Klein’s old computer that contained, to some degree, the procedures for FAEEs analysis. However, MDTL was unable to confirm when those documents were in use, if ever (they were undated), and the documents did not contain sufficient detail to enable the Independent Review to assess the adequacy and reliability of the method described.

126 In April 2008, the FAEEs procedure was MDTL’s first standard operating procedure. It became the template for the procedures that were later developed for drugs of abuse.

127 MDTL used a cut-off of 0.5 ng/mg for FAEEs, which was the cut-off established by the SoHT for the 0–3 cm sample length.MDTL would refer only those samples where there was sufficient volume of hair remaining to test for EtG.

128 The same amount of the internal standards is added to every sample to correct for any potential differences in sample preparation and extraction methods. In theory, if a laboratory uses the same robust preparation and extraction procedures for each sample, the internal standard concentrations would be the same or very similar.

Chapter 6
Adequacy and Reliability of Interpretations
and Opinions Given by MDTL

1. Introduction

1. Between 2005 and 2015, the interpretation practices of the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) and its communications with customers were inadequate, and they further undermined the reliability of the results that the Laboratory provided for use in child protection and criminal cases. It is critical that hair test results not be misinterpreted or over-interpreted and that the limitations on any interpretation be clearly conveyed to the recipient of the result.

2. MDTL generated ranges that purported to be able to identify concentrations for drugs and metabolites as “very low” (or “trace”), “low,” “medium,” “high,” and “very high.” MDTL relied heavily on these concentration ranges to interpret its test results for drugs of abuse and to communicate the meaning of those test results to its customers. Its interpretations of test results that were based on its concentration ranges were inadequate and unreliable for forensic purposes for several reasons:

  • The concentration ranges used from 2005 to 2011 were created from unconfirmed ELISA results that were preliminary in nature and unreliable both qualitatively and quantitatively. The concentration ranges were as unreliable as the underlying data.
  • Until August 2011, the concentration ranges were created from data that pooled washed and unwashed hair samples, which rendered them inherently unreliable.
  • The ranges failed to take into account inter-subject variation.
  • MDTL used the ranges improperly to reach opinions on whether a person had used drugs, in what amount, and how frequently (referred to as over-interpretation).

3. Effective communication is an important element of forensic toxicology. Forensic toxicologists are charged with ensuring not only that their tests yield accurate and reliable results but also that the results are communicated effectively to users of the results, including the justice system. MDTL inadequately communicated its test results because

  • No one at MDTL had the expertise to provide a forensic toxicology interpretation.
  • MDTL did not interpret the test result in all cases.
  • When MDTL did interpret the test results, staff often provided the interpretation over the telephone and not in writing.
  • MDTL did not communicate the limitations of its ELISA testing or the effect of the change in test methodology that began in 2010.
  • MDTL made particular interpretation errors when reporting on its tests for cocaine (and its metabolites) and cannabis / THC, two of the most common drugs for which the Laboratory tested.

2. MDTL’s Use of Concentration Ranges

4. For the reasons set out below, the interpretations that MDTL provided on the basis of its concentration ranges were not adequate and reliable for forensic purposes.

2.1 What Is a Concentration Range?

5. Starting sometime before 2005, MDTL created a series of one-page documents that grouped historical test results into ranges (originally labelled as “trace,” “low,” “medium,” “high,” and “very high”) for adults and infants for cocaine, opiates, and cannabis. The ranges were created from hair test results that MDTL had accumulated over the years. These documents, which were called “concentration ranges,” were used to interpret the results of the Laboratory’s hair tests. The earliest version of the document is set out as Figure 6.1 (and is reproduced in Appendix 8a).

Figure 6.1 Earliest version of concentration ranges for drugs of abuse provided by MDTL (in use in early 2005, if not earlier)

Figure 6.1 Earliest version of concentration ranges for drugs of abuse provided by MDTL (in use in early 2005, if not earlier)

6. MDTL provided three versions of the concentration ranges to the Independent Review:

  • 2005 (adults and infants). This earliest version of the concentration ranges did not indicate how many samples formed part of the data set and provided no information or details about how the ranges were defined. Apart from a footnote for infant samples stating that “infant hair analysis provides information regarding exposure during the last three months of pregnancy,” the document included no information about how to interpret the hair test results using the concentration ranges.
  • March 2006 (adults and infants). This version, which covered additional drugs and metabolites, used different defined terms and percentile rankings and identified the number of samples used to create the ranges for each drug or metabolite. However, it provided little information regarding interpretation of hair results.129
  • August 2011 (adult and child). The ranges in this version were created from the Laboratory’s gas chromatography–mass spectrometry (GC-MS) results and used slightly different ranges and percentiles when compared with the March 2006 ranges. These ranges specified that they were applicable only to results generated for samples received after September 1, 2010. These ranges were used until the Laboratory shut down in 2015.

7. Copies of these versions of the concentration ranges are included in Appendices 8a–8e.

2.2 Why MDTL Created Concentration Ranges

8. Julia Klein, MDTL’s laboratory manager from June 1995 to April 2005, told the Independent Review that the Laboratory created the ranges around 2002, after child protection workers asked for help interpreting the test results because the drug concentrations that MDTL reported did not otherwise have any meaning for them.

9. The Independent Review received conflicting information about who came up with the idea of using concentration ranges in this way. Ms. Klein’s recollection was that it was Dr. Gideon Koren, MDTL’s laboratory director, who suggested grouping the drug concentrations in MDTL’s database “in low, medium and high” as a way of interpreting the test results. However, Dr. Koren told the Independent Review that “[i]t was Ms. Klein’s idea to use the reference ranges.”

10. Regardless of who came up with the idea, MDTL first adopted concentration ranges in or around 2002, and the practice continued throughout the period under review, 2005–15.

2.3 Concentration Ranges Should Not Be Used Forensically

11. The concentration ranges that MDTL employed from January 2005 to August 2011 were inadequate and unreliable for use in child protection and criminal cases for two fundamental reasons: They were built on flawed data; and they failed to take inter-subject variation into account. The concentration ranges should not have been used to provide forensic toxicology opinions.

2.3.1 They Were Built on Flawed Data

12. The ranges were created from historical data from the MDTL hair-testing database. The test result data underlying the concentration ranges were deeply flawed for the following reasons:

  • The data consisted of unconfirmed ELISA (enzyme-linked immunosorbent assay) results, which are preliminary in nature and cannot be used to identify positively if a drug or metabolite is present, let alone to quantify the concentration of the drug or metabolite detected in the sample.
  • The ELISA tests had a broad range of cross-activity, particularly in the tests for opiates, amphetamine, and methamphetamine.130 Given the variability, a sample containing morphine cannot be compared reliably with a sample containing codeine (or vice versa). However, MDTL drew such inapt comparisons when it pooled together all its unconfirmed ELISA results for opiates and developed ranges to reflect its “very low,” “low,” “medium,” “high,” and “very high” results.131
  • Until September 2010, MDTL washed some, but not all, hair samples before analysis. Accordingly, the concentration ranges were created from data that pooled together results from both washed and unwashed samples (though probably mostly unwashed). MDTL’s own data indicated that washing could reduce the drug concentration in a sample by as much as a factor of four. Given this level of variation, concentration ranges created from unwashed samples (or from a combination of washed and unwashed samples) cannot provide reliable information about the extent of drug use or exposure in a particular case.
  • Given the Laboratory’s failure to standardize its sample preparation procedures, the possible changes to the extraction procedures employed, and the errors in data input identified in the sample files, the data in the database are not reliable.
2.3.2 They Did Not Account for Inter-subject Variation

13. In addition, the manner in which MDTL used its concentration ranges was inappropriate for forensic purposes owing to the significance of inter-subject variation and the many factors that can affect a hair test result.

14. Drug concentrations in hair are affected by a number of factors ranging from natural hair pigmentation and individual metabolism to cosmetic treatments and hair damage. As a result of what has been called inter-subject variation, two people who use the same amount of a drug would not be expected to have the same concentration of that drug in their hair.

15. For example, a sample may have come from a black-haired person, but the population used to create the concentration ranges may have comprised mostly blond-haired individuals. In such circumstances, the sample may have placed in the “high” range not because the person used or was exposed to more drugs than 80 percent of the Laboratory’s sample population, but rather because the population was skewed toward lighter-haired samples.132 In contrast, a sample may fall within the 10th percentile (the low range) of all samples tested by the Laboratory because the person’s hair was bleached before analysis, a process that would have decreased the concentration of the drug in the hair. Even putting aside natural hair colour and the impact of chemical hair treatments, and as discussed at length in Chapter 3, individual metabolism can have a material impact on the drug concentration that is ultimately found in a hair sample.

16. Concentration ranges increase the chance of a test result being misinterpreted or over-interpreted. It would be easy for someone to conclude that a person who tests in the “high” range used or was exposed to more drugs than a person who tests in the “low” range. Such a conclusion would be incorrect. The fact that person A may have a higher drug concentration in her hair than person B does not necessarily indicate that A used or was exposed to more of the drug than B.133

17. As a result of the significance of inter-subject variation, in the absence of sufficient information or background to provide context to a hair test result, no meaningful conclusions can be drawn from the fact that a particular result places in the very low, low, medium, high, or very high range of results generated by the Laboratory.

18. Concentration ranges should not be used to interpret hair test results for forensic purposes.134 Any information that may be gleaned from using concentration ranges derived from a laboratory’s database of results is marginal when compared with the significant likelihood of misinterpretation or over-interpretation of the result in a forensic context.135

2.4 The Laboratory Was Aware of the Pitfalls of Concentration Ranges

19. MDTL’s leadership appeared to be aware of the potential pitfalls of using concentration ranges to interpret hair test results. However, the Laboratory practices made clear that it used concentration ranges to draw inappropriate and incorrect conclusions about level and frequency of use.

20. Several documents that MDTL provided to the Independent Review noted the significance of inter-subject variation for users of the hair test results. For example, MDTL included the following caveat in a document entitled “The Use of Hair Testing to Establish Illicit Drug Use” (see Appendix 9):

However, it should be firmly stated, that the concentration of a drug in hair cannot be extrapolated to the exact amount consumed by an individual; there is a tremendous variability in the way individuals accumulate drugs in their hair, depending on the nature of the hair (dry versus oily), the colour of the hair (dark hair accumulates more on the account of higher melanin concentration), the purity of the drug and of course, the amount of drugs. With all these caveats there is one certainty: within the same individual, the more drugs the person uses or is passively exposed to, the higher the concentration in different segments of the hair.

21. Similarly, starting sometime in 2010, MDTL began to use an article that Joey Gareri (MDTL’s laboratory manager from 2005 onward) and Dr. Koren had written (together with MDTL’s first laboratory counsellor, who was a graduate student at the time) entitled “A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers” (found in Appendix 10). That article included a similar caveat:

The most commonly asked question by social workers when assessing drug test results is, “how much drug did this person use?” This question is essentially unanswerable by any existing toxicological method. Due to inherent variability in the absorption, distribution, metabolism, and elimination of drugs, subject not only to the unique characteristics of the drugs themselves, but also the user (physiological and genetic status), it is impossible to ascertain exact doses taken. Co-administration of other substances may further alter the distribution and metabolism of the drug in question.136

22. Accordingly, it appears that there were instances in which MDTL advised customers about the general limitations of hair testing. The problem, however, is that the two documents referred to above were not provided to customers as a matter of course or referred to in the results reports.

23. Mr. Gareri also told the Independent Review that MDTL chose not to provide copies of the concentration ranges to child protection workers as a matter of course to avoid having its customers attempt to interpret the results themselves. Instead, MDTL wanted its customers to call the Laboratory to obtain an interpretation in the context of a specific case.137

2.5 Despite Pitfalls, MDTL Used Concentration Ranges Inappropriately

24. However, MDTL did provide copies of its concentration ranges in a newsletter that it sent to child protection agencies, as well as in many presentations that it gave to child protection workers, hospitals, policing organizations, legal organizations, and the courts. In newsletters published in December 2005 and December 2006 (which included the concentration ranges), MDTL correctly informed customers that a hair test cannot determine the exact dose of the drug taken, and that a dose–response relationship has not been established across individuals. Nevertheless, MDTL also included the concentration ranges and suggested that meaningful information could still be obtained from these ranges.

25. Further undermining any of the caveats provided, MDTL did not convey adequately the limitations of its hair test results in the case-specific interpretations that it provided in its telephone calls and reports. To the contrary, in those case-specific interpretations, MDTL routinely relied on its concentration ranges to give opinions about the level and frequency of a person’s drug use, using such terms as “repeated,” “isolated,” or “frequent” to describe how often a person used a drug; and “bingeing” or “intensive use” to describe the quantity.138 For example:

  • One template interpretation report equated a result in the “high” range to mean “high-frequency use within the TIME PERIOD prior to hair sampling, or less frequent high-dose ‘bingeing’ within that time frame.”139
  • The 2010 interpretation guidelines stated

    LOW-range → Infrequent, isolated or sporadic use

    MEDIUM-range → Repeated use of cocaine

    HIGH-range → Frequent use of cocaine, or less frequent higher-intensity bingeing

    VERY HIGH-range → Frequent, intensive use of cocaine, likely daily

26. Accordingly, it is clear that MDTL used its concentration ranges to draw conclusions regarding the level and frequency of use by an adult not only in its internal documents but also in its communications to users of the test results. Such a practice was not acceptable for a laboratory providing forensic services for the following reasons:

  • MDTL advised that the results indicated use of drugs rather than exposure to drugs. Such a conclusion cannot be determined simply by placing a result within a set of concentration ranges and is particularly inappropriate where samples have not been washed before testing.
  • Because of inter-subject variation, concentration ranges cannot be used with a hair result to determine the amount of drugs that a person has consumed or how often the individual has used the drug within the time frame being tested.

27. It was overly simplistic, and misleading, for MDTL to reach conclusions about the frequency and intensity of a person’s drug use based solely on where a result placed in the Laboratory’s concentration ranges. Similar problems were caused by MDTL’s use of concentration ranges in child samples. Although MDTL was more explicit about the need to consider external contamination when interpreting child samples,140 it continued to use concentration ranges for cocaine and benzoylecgonine for its unwashed samples (for children) to draw inferences about how the drug or metabolite may have got into the child’s hair.

28. When an unwashed hair sample tests positive for a drug, all that can be said is that the drug was detected. No further inference can be made about whether the drug was on the surface of the hair or in it. Because external contamination can occur in many ways and the concentrations can vary dramatically depending on how the drug got into the hair, very little can be gleaned from an unwashed hair sample. It was inappropriate for MDTL to use the results from unwashed samples to assess the frequency of a caregiver’s use or the level of risk in the child’s environment in reference to the Laboratory’s concentration ranges. Without more information about the child’s environment or individual circumstances, MDTL’s interpretation should not have gone any further than to say that the drug was detected.

29. The interpretation reports that the Independent Review randomly selected and reviewed for child samples also included statements to the effect that “caregiver drug use is occurring within the context of caring for the child.” Although it may be appropriate to infer that the drugs or metabolites that were found in a hair sample may have originated from a caregiver (e.g., as a result of frequent handling of the child), it is inappropriate to infer from the hair test result that a caregiver was using drugs around a child at the time of the transfer. A hair test cannot determine when the drug or metabolite got into the child’s hair, including if it was during a time in which a caregiver was fulfilling a parenting role.

30. Accordingly, although MDTL appeared to appreciate the unique factors that need to be considered when interpreting child hair test results to some degree, including the increased risk of external contamination, MDTL also over-interpreted its hair test results for child samples.

2.6 Concentration Ranges in Use: September 2010–15

31. After 2010, MDTL stopped using unconfirmed ELISA tests for many, but not all, of the drugs of abuse for which it tested. MDTL’s continued use of concentration ranges for drugs of abuse in the post–September 2010 period remained problematic for a number of reasons.

32. First, Mr. Gareri told the Independent Review that, from September 2010 until August 2011 (when MDTL released its new set of concentration ranges), the Laboratory interpreted its GC-MS results using its March 2006 concentration ranges (which, as noted above, were created using MDTL’s unconfirmed ELISA results). Mr. Gareri recalled advising users that “the results might be off because there might be variation in how the instruments (GC-MS vs. ELISA) quantify the concentrations.”

33. However, MDTL’s interpretation of its GC-MS results with reference to the March 2006 ELISA concentration ranges is a serious concern. For the reasons described above, the data underlying the March 2006 concentration ranges were flawed. Interpreting a GC-MS result using those ranges is equally flawed. Furthermore, despite Mr. Gareri’s recollection, the Independent Review did not identify any documents in the randomly selected post–September 2010 case files it reviewed that communicated to users of the test results that the ranges might be “off” on account of a difference in the analytical instrument used.

34. Second, as a result of the deficiencies in MDTL’s GC-MS procedure, the data underlying the August 2011 concentration ranges may not have accurately reflected the drug concentrations in the samples that MDTL tested. Third, the use of concentration ranges continued to include results from unwashed samples. Starting in September 2010, MDTL routinely washed all adult samples before testing; however, the Laboratory chose not to wash child and neonatal samples on the theory that it wanted to detect the presence of any drugs or metabolites in or on the sample (whether as a result of active drug use or passive contamination). Accordingly, the data underlying the August 2011 concentration ranges for cocaine and benzoylecgonine in child samples continued to be based on results from samples that were not washed before testing. For the reasons set out above, no meaningful conclusions can be drawn from concentration ranges created from such data.

3. MDTL’s Communication of Its Test Results

3.1 Nobody Had the Expertise to Give Forensic Interpretations

35. The interpretation of a hair test result for forensic purposes requires a significant level of expertise. It should be performed only by a properly trained and qualified expert. The MDTL staff members who provided the interpretations did not have the necessary expertise to provide forensic interpretations.

36. Dr. Koren rarely provided interpretations to child protection workers or clinicians. He occasionally testified in court – for example, at the Broomfield trial in 2009 – and he provided an expert report on the appeal in January 2013. Dr. Koren saw himself as an expert in the interpretation of results. However, given the serious flaws in the analytical methods used in the Laboratory for which he was responsible and his continued championing of the ELISA methodology in the Broomfield appeal, it is clear that he did not understand basic elements of forensic toxicology that are necessary to provide interpretations of test results. As the laboratory director, Dr. Koren was responsible for ensuring that the individuals who were providing routine interpretations were properly trained and qualified to do so, and that they provided reliable interpretations to customers and other users of the test results in individual cases. He did not do so.

37. When Mr. Gareri was appointed laboratory manager, he had no formal training or experience in forensic toxicology, hair analysis, or the interpretation of hair test results. Over time, Mr. Gareri gained experience in clinical and forensic toxicology and the science of hair testing.141 However, he lacked the necessary training and expertise

  • on the interpretive aspects of hair analysis, including the limitations on interpretations and the many factors that should be considered when interpreting a specific result;
  • on the analytical aspects of the Laboratory’s hair tests for drugs of abuse and their limitations (including knowledge that ELISA tests were preliminary screening tests only);
  • to explain the adequacy and reliability of the test results, even if he was not capable of performing the tests personally; and
  • to understand the role of the expert witness in legal proceedings.142

38. The leadership of MDTL and the Hospital for Sick Children (SickKids or the Hospital) should neither have given Mr. Gareri the responsibility to interpret hair test results nor held him out as an expert to participants in the child protection and criminal justice systems. Nevertheless, shortly after he was appointed, he began frequently and routinely interpreting the test results for those involved in child protection and criminal proceedings.

39. By 2006, Mr. Gareri was testifying as an expert witness in court, despite having received no training from the Hospital or Dr. Koren regarding the need to be objective and to communicate any limitations on or controversies surrounding the expert’s opinion in a clear and transparent manner. Dr. Koren was not trained to testify as an expert and could not have provided Mr. Gareri with appropriate training. If the Hospital did not have someone capable of providing this training, it should have obtained the necessary expertise. Mr. Gareri offered the Independent Review the following perspective about his lack of training at the time he testified at the Broomfield preliminary inquiry:

In retrospect, [I do] not think [I] was properly trained to act as an expert at the time of the preliminary inquiry (in 2007). At the time, [I] asked Dr. Koren to testify at the trial because [I] was not comfortable testifying in a criminal trial of that magnitude. In hindsight, [I believe I] should have been more conscious of the limitations of [my] expertise when giving evidence during these earlier years and in fact, should not have been going to court at all around this time due to a lack of formal training on the role of the expert.

40. Mr. Gareri was not provided with the training required to offer adequate and reliable interpretations of MDTL’s hair test results either on the stand as an expert witness or on the telephone or in writing to customers and other users of MDTL’s hair tests.

41. Starting in 2008, Mr. Gareri assigned a graduate student (whom MDTL called a “laboratory counsellor”) to assist in providing the over-the-phone consultations regarding the Laboratory’s hair test results. By 2015, 16 individuals had worked at MDTL as laboratory counsellors. With two exceptions, all 16 were at the time graduate students or fellows in pharmacology or clinical pharmacology and toxicology at the University of Toronto.143

42. MDTL promoted its use of laboratory counsellors as an efficient way to provide over-the-phone interpretations to child protection agencies. In its December 2009 “Motherisk Drug Testing Newsletter for Children’s Aid Societies,” MDTL noted that its “free consultations had snowballed into one of the major components of our service.”144

43. The title “laboratory counsellor” masked the fact that these graduate students and fellows worked only short-term and part-time for MDTL and provided no counselling function whatsoever. They provided opinions interpreting MDTL’s hair test results, which should have been given by experts in hair analysis with the training and qualifications to do so. In any event, because MDTL referred to these individuals as “laboratory counsellors,” I also use that term in this Report.

44. The laboratory counsellors were not properly trained or qualified to provide reliable interpretations of the test results in individual cases. Nevertheless, the laboratory counsellors became responsible for answering what MDTL referred to as “routine” calls for interpretations of hair test results and were instructed to refer any “non-routine” calls to Mr. Gareri. Beginning in 2011, the laboratory counsellors also drafted the majority of the interpretation reports that MDTL issued, basing them on previously issued reports and templates that Mr. Gareri had prepared. However, they did not sign the reports. Nor were they called to testify in court.

45. Until 2012, the laboratory counsellors had no formal training whatsoever. Instead, the training consisted of having the graduate students or fellows sit in on consultation calls with Mr. Gareri, to listen to how he answered questions from MDTL’s customers. MDTL also gave the graduate students and fellows a binder of sample interpretation reports and other material relevant to the interpretation of hair test results.145 In November 2009, the Laboratory developed its interpretation guidelines and provided them to the laboratory counsellors to use as a “script” when speaking to customers.146

46. In 2012, MDTL developed a more formalized training program for its laboratory counsellors, which included five one- to two-hour presentations on the principles of hair analysis generally, as well as neonatal and pediatric hair analysis specifically. The laboratory counsellors also continued to receive reference materials and “shadowed” the more experienced laboratory counsellors by sitting in on calls before answering any themselves.

47. This training was woefully inadequate. The laboratory counsellors lacked the necessary training and expertise

  • on the interpretive aspects of hair analysis, including the limitations on interpretations and the many factors that should be considered when interpreting a specific result;
  • on the analytical aspects of the Laboratory’s hair tests for drugs of abuse and their limitations, including that ELISA tests were preliminary screening tests only; and
  • to explain the adequacy and reliability of the test results.

48. As set out above, an opinion interpreting a hair test result is an expert opinion, which should be provided only by an individual with appropriate training and qualifications. The graduate students and fellows MDTL employed to interpret hair test results – who had no experience in the analytical aspects of the tests on which their opinions were being sought and no formal training in forensic toxicology or the science of hair testing – were not experts in hair analysis. They lacked the training and expertise to provide adequate and reliable opinions for use in child protection and criminal cases.

3.2 MDTL’s Interpretations Not Appropriate

3.2.1 MDTL Did Not Provide an Interpretation in Each Case

49. Hair test results, viewed in isolation, cannot provide any meaningful information about drug use or alcohol consumption and can be potentially misleading. To provide any meaningful insight into a person’s use of or exposure to drugs or alcohol, the hair test result must be interpreted in the context of the particular case, with regard for the various factors that may affect the result.

50. MDTL did not routinely provide an interpretation of the hair test results to its customers. Instead, it provided the numerical test results to customers (who may have also had the concentration ranges) and invited the customer to call an MDTL staff person or laboratory counsellor for further information. (For a sample results report, see Appendix 11a.) Providing the numerical test results, and nothing more, suggested a degree of precision and ease of interpretation that was incorrect and would have misled customers.

51. MDTL failed to communicate to the customer adequately, or at all, the limitations in hair analysis – including the lack of a dose–response relationship across individuals and the potential impact of external contamination on the interpretation of hair test results for drugs of abuse.147 Without an adequate understanding of what the hair test result can and cannot say in the circumstances of a particular case, there is a material risk of misinterpretation or over-interpretation of the results by the end-user.

52. This possibility applies equally to the Laboratory’s interpretations of alcohol markers. Although the concentration ranges that MDTL developed for alcohol markers contained more detail over time (see Appendices13a–13e) and MDTL told the Independent Review that those concentration ranges were sent out with every results report, I am still concerned that case-specific opinions were not provided as a matter of course. For example, although MDTL recognized the potential impact of ethanol-containing hair care products on fatty acid ethyl esters (FAEEs) results,148 Mr. Gareri told the Independent Review that the Laboratory requested this information only when the results were disputed. However, given the potential impact of ethanol-containing hair care products on FAEEs results, MDTL should have obtained a list of hair care products used and their ingredients (or at least stated the potential impact of such products for individuals who could not recall which ones they used) and interpreted its FAEEs results in the context of that information for every FAEEs result that it reported. In other words, it was incumbent on MDTL to provide a meaningful and reliable interpretation of the test result to its customer at the outset – the Laboratory should not have waited to see if the results were ultimately disputed before doing so.

53. Although it may be acceptable in some cases for a forensic laboratory to provide no opinion with a hair test result – for example, where the result is negative; or in the workplace drug-testing context, where the results will be considered and interpreted by a medical review officer149 – given the many factors that may affect a hair test result and the risk of misinterpretation or over-interpretation, all the cases in which MDTL reported a positive hair test result for use in child protection or criminal cases should have been accompanied by a written opinion interpreting the result within the context of the particular case.150

3.2.2 MDTL Rarely Provided Written Interpretations

54. Given the complexities in hair testing and the various factors that must be considered in the interpretation of a hair test result, opinions interpreting hair test results must be provided in writing to avoid any misunderstanding by the user. However, throughout the period from 2005 to 2015, unless the customer requested a written interpretation report, MDTL’s practice was to provide interpretations of its test results over the telephone. In fact, in comparison with the number of samples that MDTL tested, written interpretation reports were rarely prepared. Accordingly, in most cases there was no written record of the expert opinion provided to the customer in respect of the particular result.

55. Without a written expert opinion, it is difficult, if not impossible in some cases, to determine if an opinion interpreting a hair test result was provided in a particular case and, if it was, by whom and what was said.

56. Prior to 2009, staff at the Laboratory kept no records of the oral opinions provided to its customers. After 2009, note-taking was haphazard at best, despite the introduction that year of Consultation Intake forms. These forms, used to keep records of telephone consultations by MDTL’s laboratory counsellors, did not identify the person providing the interpretation, were used irregularly (for example, Mr. Gareri did not routinely use them or keep any other record of the interpretations he provided over the telephone), contained varying levels of detail, and were not provided to the caller.

57. Although it is acceptable for experts to discuss opinions of test results over the telephone, the fundamental flaw in MDTL’s practice was that, in many cases, no written opinion of any kind was provided to users of the results. The Consultation Intake forms may well have been an appropriate way to document a discussion in a case in which a written interpretation report had been provided and the caller was phoning to discuss that report. It was not, however, adequate if it was the only document recording the opinion that MDTL provided to the customer interpreting the hair test result. Given the many complexities involved in the interpretation of hair test results for drugs of abuse and alcohol markers and the need to reduce the risk of misinterpretation or over-interpretation, a written opinion interpreting the hair test results should have been provided every time.

3.3 MDTL Did Not Communicate the Limitations of ELISA

58. MDTL failed to inform users of what test or tests it carried out and of any limitations on the analysis or the result. Customers and other users of MDTL’s hair test results – among them the justice system – were not advised of the significant limitations of the ELISA test, including its inability to identify and quantify drugs or metabolites – both before and after the Laboratory changed testing methodologies in September 2010.

59. Between January 2005 and August 2010, MDTL failed to advise customers that the Laboratory used only ELISA to test its hair samples. The Laboratory also failed to identify for customers or other users of its test results the significant limitations in the ELISA test, including its inability to identify or quantify the drug or metabolite concentration in a sample. MDTL routinely reported its unconfirmed quantitative ELISA results from January 2005 to August 2010 and interpreted those results in the context of its concentration ranges, without any suggestion that the ELISA test could not be used reliably either to identify or quantify the amount of drug in the sample. Instead, MDTL reported and interpreted its unconfirmed ELISA results throughout this period as though it had positively identified and quantified the drug or metabolite in question, when in fact it had not. MDTL routinely interpreted its unconfirmed ELISA results, both qualitatively and quantitatively, without regard for the fact that those results were only preliminary in nature.

60. Some of the interpretation reports that the Independent Review received from the 2005–10 period were potentially misleading because they suggested that MDTL had carried out a “secondary analysis” when a sample was first tested as a whole (or in bulk) and then retested segmentally. For example, an interpretation report for a sample collected in August 2010 that MDTL initially tested in bulk and then segmentally (both using ELISA testing) stated: “The initial positive opiate findings were not confirmed upon secondary analysis of this sample; indicating that opiate use by this individual is not conclusively determined by the results of this hair test.”

61. The suggestion that segmental analysis using ELISA was a “secondary analysis” that could “confirm” the initial finding was misleading because ELISA is a preliminary screening method. Until the ELISA results are confirmed by a confirmation test such as GC-MS or LC-MS/MS, they should not be reported as positive. A second ELISA test run on the same sample is not a way to confirm the initial result. MDTL should have made clear to the user that both tests were preliminary in nature. Without a confirmation test, MDTL should not have said anything more than that.

62. Dr. Koren advised that he was aware that ELISA could not generate quantitative results as reliably as a confirmation test, such as GC-MS. He told the Independent Review that, as a result, when he testified in court about ELISA results, he “focused on the qualitative result in question.” However, the fact that MDTL used concentration ranges and routinely interpreted the results with reference to those ranges indicates that the Laboratory considered ELISA to be a quantitative test.

63. In addition, Dr. Koren testified at the Broomfield trial in 2009 about the Laboratory’s ELISA test results. That testimony did not focus on the qualitative result. During his examination-in-chief, Dr. Koren testified about the quantitative result by describing the “high” or “very high” levels of cocaine and benzoylecgonine found in the child’s hair, and he went so far as to say that “these high levels [found in the child’s hair] in an adult represent severe addiction and drug dependence” and “[t]hese levels are the highest our laboratory ever measured.”

3.4 MDTL Did Not Explain the Change in Testing Methodology

64. In September 2010, MDTL transitioned to routine GC-MS confirmation for the majority of the drugs and metabolites that it tested in adult and child samples. Mr. Gareri acknowledged to the Independent Review that “customers did not understand the difference between immunoassay and GC-MS testing.” He told the Independent Review that he “did not get into the methodological minutiae when he communicated with customers – he just focused on the interpretation of the end result.” Unfortunately, the Laboratory did not explain what that change meant for the customers and other users of its hair test results.

65. On October 15, 2010, MDTL sent a customer notification memorandum to its frequent customers, including child protection agencies, about “several exciting new service improvements” at the Laboratory (see Appendix 14). The memorandum was misleading in several ways:

  • It did not disclose the fact that MDTL was moving from a preliminary screening test (ELISA) to routine GC-MS testing for the majority of the drugs and metabolites that it tested in adult and child samples.
  • It did not state that the ELISA tests that MDTL had previously used were preliminary and unconfirmed; it also failed to state that the qualitative results that it began to report for neonatal samples and for the drugs for which it did not have GC-MS capability were preliminary and unconfirmed.
  • It suggested that the Laboratory was reporting cannabis / THC and other drugs qualitatively rather than quantitatively because qualitative reporting “will have no negative impact on the interpretation of results for these compounds and allow for a quicker turnaround of results.” The move to qualitative reporting for the drugs and metabolites for which MDTL had not developed an in-house GC-MS method resulted from recognizing that ELISA was unreliable for quantitative purposes.

66. MDTL also did not explain adequately its testing methodologies in the interpretation reports that it provided to its customers. As a result, the Laboratory created the risk of conflating the preliminary screening test with the GC-MS–confirmed results. For example, after September 2010

  • the Laboratory’s results reports and interpretation reports used the term “screen” to refer to ELISA results for opiate and amphetamine samples that were then tested using GC-MS151 but not in respect of samples that were tested using ELISA only, such as cannabis / THC and benzodiazepines;
  • in some cases, the Laboratory used “secondary analysis,” without explaining the type of analysis used (GC-MS or a second ELISA test); and
  • it failed to indicate that the results for the drugs and metabolites that were tested only with ELISA were not confirmed by any method.

67. After MDTL transitioned to LC-MS/MS testing in May 2014, its interpretations were more consistent in the identification of the testing methodology used. It appears, however, that although MDTL had stopped using ELISA to screen for opiates (as well as for amphetamines and methamphetamines) in May 2014, it continued to use the same template, which indicated that a “screen” had been employed. MDTL’s continued use of the “screen” language in the context of its LC-MS/MS tests was confusing at best, and misleading at worst.

3.5 Issues Raised by Interpretations for Cocaine and Cannabis / THC

68. The Independent Review identified several issues with MDTL’s interpretations for cocaine (and its metabolite) and cannabis / THC, which were two of the drugs and metabolites that were most frequently reported as positive and are therefore among the most important of the tests. Although they improved over time, the interpretations were inadequate for forensic purposes in the following ways.

69. First, Dr. Koren stated that, from 2005 to August 2010, MDTL reported a positive result only when the test for cocaine was positive and the test for benzoylecgonine yielded a result that was greater than 5 percent of the cocaine concentration. The 5 percent rule to which Dr. Koren referred comes from the Society of Hair Testing’s (SoHT’s) 1997 consensus statement for drugs of abuse, which recommended use of a benzoylecgonine-to-­cocaine ratio of 5 percent to distinguish between active use of and passive exposure to cocaine. Although a 5 percent metabolite-to-drug ratio was the consensus approach in 1997, later studies refuted the reliability of relying on only the metabolite-to-drug ratio, and the SoHT’s later statements on the examination of drugs of abuse in hair (dated 2004 and 2012) no longer included metabolite-to-drug ratios for the interpretation of hair test results for drugs of abuse. Moreover, those later SoHT consensus statements make it clear that the role of external contamination must be considered in every case involving hair testing for drugs of abuse. Accordingly, MDTL should not have relied on the 5 percent benzoylecgonine-to-cocaine ratio as a hard-and-fast rule, and it should have highlighted for the reader the possibility of external contamination, notwithstanding the detection of benzoylecgonine.

70. Second, from 2005 to August 2010, MDTL relied on the presence of trace amounts of benzoylecgonine to confirm active ingestion of cocaine. However, the suggestion that a trace amount of benzoylecgonine is indicative of active use of cocaine is an over-interpretation of the result. MDTL should not have relied on a trace amount of benzoylecgonine to reach such a conclusion. Putting aside the flaws in interpreting the quantitative results from unwashed samples using ELISA, the fact that benzoylecgonine may be present in trace quantities is not necessarily indicative of active cocaine use. In fact, benzoylecgonine can be created by hydrolysis (including during the extraction process in the Laboratory). MDTL did not account for this possibility in its interpretations.

71. Finally, from 2005 to August 2010, MDTL relied on ELISA results of cannabis / THC tests on (likely) unwashed samples to reach conclusions on the frequency of a person’s use of cannabis. For example, MDTL’s template interpretation report, which was last modified in May 2007, provided:

Cannabinoids: The negative result for cannabinoids indicates no evidence of cannabis use in this individual during the time period tested. It should be noted that due to certain physical and chemical properties, cannabinoids are not taken up in hair to the same degree as cocaine. The concentrations of cannabinoids that are found in hair are much lower and individuals generally must use cannabis several times a week in order to test positive in hair. [Emphasis added.]

72. Although MDTL was correct to recognize that cannabis / THC, which is an acidic drug, does not incorporate into hair as readily as basic drugs (such as cocaine and certain opiates and amphetamines), the conclusions that MDTL reached during this period overstated what the results could say. Cannabis / THC is a drug that can be smoked and is therefore susceptible to external contamination. Even putting aside MDTL’s reliance on ELISA tests on unwashed samples (which, for the reasons set out above, is a significant issue), there was no scientific basis to suggest that a cannabis / THC finding could, on its own, indicate the frequency of a person’s drug use. MDTL also did not give sufficient weight, and in some cases any weight, to the possibility of external contamination.


Notes

129 Both the earlier version and the March 2006 version of the concentration ranges were created from MDTL’s database of preliminary and unconfirmed ELISA results. The Independent Review was not provided with details of the data used to create the earlier version of MDTL’s concentration ranges; however, MDTL told the Independent Review that the March 2006 concentration ranges were created using all the available positive results from MDTL’s database from 1998 to September 2005.

130 As described in Chapter 5, ELISA tests may cross-react with compounds that are structurally related to the target drug or metabolite, and the ELISA tests for different drugs and metabolites have varying degrees of cross-reactivity. ELISA may also show an apparent positive response when no drug is present owing to the presence of other substances or interferents.

131 The ELISA test for amphetamine detected both amphetamine and MDA, and the ELISA test for methamphetamine detected both methamphetamine and MDMA. Accordingly, the concentration ranges that MDTL developed for its unconfirmed ELISA results for amphetamine and methamphetamine suffered from the same deficiency.

132 As explained in Chapter 3, the difference in detecting certain drugs among individuals with different hair pigmentation is referred to as “hair colour bias.”

133 As explained in Chapter 3, although controlled dosage studies have supported an approximately linear relationship between dose and drug concentrations within an individual, there is not a similar linear relationship between dose and response when considering hair test results from different people – likely because of inter-subject variation.

134 At least one hair-testing expert has proposed using statistics to develop ranges of low, medium, and high concentrations (similar to how MDTL created its concentration ranges) as a guideline to interpret individual hair test results. See C. Jurado, “Forensic Applications of Hair Analysis,” in P. Kintz, A. Salomone, and M. Vincenti, eds, Hair Analysis in Clinical and Forensic Toxicology (Amsterdam: Elsevier, 2015) 241–73, which refers to population studies carried out in Switzerland, the United Kingdom, and Spain using low, medium, and high ranges. The experts retained by the Independent Review, however, were not aware of any hair-testing laboratories other than MDTL that use concentration ranges as an interpretive tool.

135 I note that, in the clinical setting, reference ranges or intervals are commonly used to express the range of values for a physiological measurement to distinguish between normal levels and those that might suggest disease. The use of reference ranges in that context is a widely accepted practice. I make no comment on the use of reference intervals for diagnostic, prophylaxis, or treatment purposes; my conclusions in respect of MDTL’s concentration ranges relate to the forensic purposes for which they were used.

136 Monique Moller, Joey Gareri, and Gideon Koren, “A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers” (2010) 17(1) Canadian Journal of Clinical Pharmacology e177–e93 at e179.

137 The August 2011 concentration ranges (see Appendices 8d and 8e) also stated: “These ranges are a guideline regarding average intensity of drug use over time; toxicology results should be interpreted by a qualified expert.”

138 MDTL provided a random sampling of interpretation reports from 2005 to 2015. As well, MDTL created template interpretation reports and guidelines that it also provided to the Independent Review. The template reports contained MDTL’s standard written interpretations for the more commonly tested drugs of abuse, and the interpretation guidelines were provided to laboratory counsellors to use in consultation calls with customers and other users of the hair test results. Copies of three template interpretation reports and the interpretation guidelines MDTL produced to the Independent Review are found in Appendices 11b, 11c, 11d, 12a, and 12b.

139 This example is from a template interpretation report that was last modified in May 2007 for adult samples.

140 See, for example, the August 2011 concentration range document in Appendix 8e.

141 Mr. Gareri joined the Society of Hair Testing and attended its annual conferences. He also attended conferences of other international organizations, including the International Association of Therapeutic Drug Monitoring and Clinical Toxicology (IATDMCT) and the International Association of Forensic Toxicologists (TIAFT).

142 For example, the Forensic Toxicology Laboratory Guidelines of the United Kingdom and Ireland Association of Forensic Toxicologists (UKIAFT), which were published in 2010, specifically recognize the need for such training for the individuals who are involved in the reporting of test results: “4.2.2. One or more scientists who supervise the work of all technician / analysts, and should be capable of carrying out [a] full scientific review of all test data. Acceptable qualifications include a minimum of a Bachelor’s degree in one of the natural sciences or recognised equivalent. They should be trained and deemed competent in all appropriate methods and procedures before reporting and acting as expert witnesses. All reporting scientists should receive expert witness training.”

143 One of the laboratory counsellors was a post-doctoral fellow at SickKids’s Research Institute and another had previously worked as a manager / technical associate in toxicology at a forensic toxicology laboratory in France.

144 MDTL charged customers for the hair test but did not charge for interpretations, including the over-the-phone interpretations, the preparation of interpretation reports, or court time.

145 Starting in 2010, laboratory counsellors also received for use as a reference a copy of the article that Dr. Koren and Mr. Gareri (together with MDTL’s first laboratory counsellor) had published in Canadian Journal of Clinical Pharmacology, “A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers” (see Appendix 10).

146 The interpretation guidelines produced to the Independent Review are found in Appendix 12.

147 As noted, some of the documents that MDTL provided to the Independent Review, including the document entitled “The Use of Hair Testing to Establish Illicit Drug Use” (see Appendix 9) and the article that Dr. Koren and Mr. Gareri had authored with MDTL’s first laboratory counsellor (see Appendix 10), contained an important caveat about the significance of inter-subject variation and the inability to determine from a drug concentration the exact amount of the drug consumed by the individual. However, those documents were not sent to customers as a matter of course – and they were provided only when MDTL’s test results were being submitted for court or if a customer requested general information about the capabilities of hair analysis. Accordingly, I am not satisfied that most customers would have received the information in these documents about the uses and limitations of hair analysis.

148 FAEEs comprise a group of more than 20 minor metabolites of ethanol and are one of the two main alcohol markers currently tested in hair. Starting in July 2007, the FAEEs concentration ranges began to indicate that low concentrations of FAEEs may be found in the hair of non-drinkers owing to “the use of certain hair-care products.” In March 2010 the FAEEs concentration ranges sheet began to indicate more explicitly that high ethanol-containing hair care products can lead to higher concentrations of FAEEs.

149 A medical review officer is a person, usually a physician, who is responsible for reviewing drug test results generated by a workplace drug-testing program.

150 The 2006 version of Forensic Toxicology Laboratory Guidelines of the Society of Forensic Toxicologists and American Academy of Forensic Sciences (SOFT/AAFS Guidelines) actually discourages comments in toxicology reports. However, I understand from the experts retained by the Independent Review that this approach is largely confined to the United States, where post-mortem toxicology results are routinely given to the forensic pathologist to consider with all the other findings in the case. In any event, I prefer the Forensic Toxicology Laboratory Guidelines of the United Kingdom and Ireland Association of Forensic Toxicologists (UKIAFT Guidelines), which were published in 2010 and call for the interpretation of toxicology results where adequate background information is available.

151 As described in Chapter 5, the ELISA kit for opiates cross-reacted with morphine, 6-AM, and codeine, as well as a host of other prescription and non-prescription drugs. Starting in approximately September 2010, MDTL also stopped using the term “opiates,” opting to use the term “opioids.” Opiates refer to morphine and drugs derived from morphine, whereas opioids refer to all drugs acting like morphine on the opioid receptor systems (such as methadone, meperidine, or fentanyl). I do not consider that much turns on the distinction, and I use the term “opiates” in this Report. In addition, although MDTL used two ELISA kits to screen for amphetamine (an ELISA kit for amphetamine, which cross-reacted with MDA, and an ELISA kit for methamphetamine, which cross-reacted with MDMA), the Laboratory referred to the amphetamines as comprising a single “screen.”

Chapter 7
Proficiency Testing and Accreditation

1. Introduction

1. There are two external measures for assessing the adequacy and reliability of a laboratory’s analytical tests and operations:

  1. proficiency testing, which is an inter-laboratory comparison used to assess how a laboratory’s test results fare when compared with other laboratories carrying out similar tests; and
  2. laboratory accreditation, which involves an assessment by a third party (an accrediting body) of the laboratory’s procedures.

The two concepts are not mutually exclusive, as the former can be (and often is) a requirement of the latter. Although neither one is required for a laboratory to carry out forensic tests in Ontario, a laboratory’s performance in proficiency tests and/or the fact it has been accredited by a third party can serve as an indicator of the reliability of the laboratory’s tests.

2. When questions were raised about the use of its enzyme-linked immunosorbent assay (ELISA) hair tests on and following the Broomfield appeal, the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) emphasized its participation in an internationally recognized proficiency-testing program offered by the Society of Hair Testing (SoHT) and its participation in the Hospital for Sick Children’s (SickKids’s or the Hospital’s) clinical laboratory accreditation process to support the scientific reliability of its hair tests for drugs of abuse. However, as I describe in this chapter, neither MDTL’s proficiency test results nor its accreditation gives me any additional confidence in the adequacy or reliability of MDTL’s hair tests during the period under review.

2. MDTL and Proficiency Testing

3. Proficiency testing is an external assessment of the quality of a laboratory’s results by means of an inter-laboratory comparison. From 2001 to 2014, MDTL participated in proficiency testing for drugs of abuse offered by the SoHT. From 2011 to 2014, MDTL also participated in the proficiency-testing scheme that the SoHT offered for fatty acid ethyl esters (FAEEs).

4. For its proficiency tests, the SoHT distributes three to five hair samples to participating laboratories each year.152 Because the SoHT uses real hair samples, as opposed to hair that has been fortified with a known concentration of a drug or metabolite, it uses reference laboratories to quantify the amount of the drug or metabolite in each of the hair samples. The samples that the SoHT sends to participating laboratories are “blinded” in the sense that the participating laboratories do not know whether the samples contain any drugs or metabolites. The participating laboratories are supposed to test the blinded samples using their standard procedures to determine if the samples contain drugs or metabolites, and if they do, what they are, and in what concentrations.

5. For example, in 2001 (the first year in which MDTL participated in the SoHT proficiency-testing scheme), the SoHT’s proficiency test involved 18 laboratories from the United Kingdom, other parts of Europe, and North America, including MDTL. The SoHT distributed three hair samples to the laboratories. The first sample contained amphetamines and cannabis, the second contained opiate and cocaine compounds, and the third was drug-free. The SoHT requested that participating laboratories “analyse the samples following the standard procedure used routinely in your laboratory as closely as possible.” On the proficiency test submission forms, laboratories were asked to include information about their analytical methods and to identify the instruments used both as a screening test and for “confirmation and quantification.” If the laboratory offers both screening and confirmation, it is then required to submit two results for each sample:

  • the result obtained from the laboratory’s screening test (reported qualitatively); and
  • if the sample screened positive, the laboratory’s quantitative result for the drugs and/or metabolites identified in the sample, obtained from its confirmation test.153

6. The proficiency test submissions followed this format each year. Once the test was completed, the SoHT provided each participating laboratory with a report detailing how the laboratory fared in comparison to the SoHT’s reference laboratories, as well as to the other participating laboratories.

7. Because MDTL’s proficiency tests raise different issues for drugs of abuse in the pre-2010 period, drugs of abuse in the post-2010 period, and alcohol markers, I address each category separately.

2.1 Drugs of Abuse: Pre-2010

8. MDTL used ELISA to test for drugs of abuse without routine confirmation of the results until August 2010. When questions were raised about the reliability of its ELISA tests both on the Broomfield appeal and in its aftermath, MDTL relied on its performance in the SoHT’s proficiency-testing scheme from 2001 to 2009 to support the reliability of its ELISA test results:

  • On the Broomfield appeal, MDTL’s laboratory manager, Joey Gareri, prepared a report, dated January 18, 2013, responding to the concerns raised by the defence’s expert, Dr. Craig Chatterton. In his report, Mr. Gareri maintained that MDTL’s proficiency-testing history established that the Laboratory’s ELISA results were comparable to GC-MS–based results:

    Regarding our proficiency testing, we have participated in the international proficiency testing program provided by the Society of Hair Testing (SOHT) for over ten years. The international SOHT program is the only available proficiency testing program for hair analysis of drugs in Canada. To address Dr. Chatterton’s concern regarding our proficiency testing performance, I have included in my report our proficiency testing results from 2005 ... You will note that our laboratory’s performance for cocaine analyses show[s] our test methods to be within one standard deviation of the mean of participating and reference laboratories and that our benzoylecgonine analyses demonstrate that our laboratory was showing a bias towards underestimating benzoylecgonine levels in hair. In our entire proficiency testing history during my tenure managing this laboratory (2005 to present), we have no instances of false-positive or false-negative findings for cocaine or benzoylecgonine. [Emphasis added.]

  • Similarly, in response to Dr. Chatterton’s criticism that immunoassays, such as ELISA, do not yield accurate quantitative results, Mr. Gareri wrote in a supplementary report, dated June 17, 2013:

    Since most immunoassays exhibit substantial cross-reactivities against metabolites; it is often true that immunoassays cannot yield accurate quantitative results. Regarding the results of interest in this case; I can state from our proficiency testing history that our cocaine and benzoylecgonine immunoassays have produce[d] comparable results to those in MS-based labs. [Emphasis added.]

  • On the Broomfield appeal, MDTL’s laboratory director, Dr. Gideon Koren, also prepared a report, dated January 11, 2013, in which he supported the accuracy of MDTL’s quantitative ELISA tests by describing MDTL’s proficiency test results for cocaine and benzoylecgonine in 2005 (the year in which the child in Broomfield’s sample was tested) as “on the spot”:

    The Society established the first ever quality assurance program, whereby all participants receive unknown hair samples and measure in them drugs [sic]. This has been the only international process to assess the quality of results in laboratories across the world. MDTL has participated in this process since its inception, and our results have been consistently in agreement with the Round Robin consensus. Dr. Chatterton acknowledges this process in his report, but claims that we did not show specific results. Tab 1 presents our performance in this test for the year of [the child’s] case, ie 2005. It shows very clearly that our measurements of cocaine and its metabolite [benzoylecgonine] were right on the spot. This further addresses Dr. Chatterton’s criticism that we did not conduct GC-MS but rather “radioimmunoassay” test. Dr. Chatterton wrongly states that we conducted radioimmunoassay; rather, we performed ELISA assays. The fact that our ELISA results are similar to those achieved by GC-MS addresses this question. Between 2005 and now, we did not have a case of false positive or false positive [sic] results of Cocaine or [benzoylecgonine]. [Emphasis added.]

  • In November 2014, following the Court of Appeal’s decision in Broomfield, MDTL prepared an information sheet for distribution to child protection agencies by the Ontario Association of Children’s Aid Societies. Again, MDTL highlighted that it had “very good performance” in the SoHT’s performance-testing scheme since 2001:

    The vast majority of laboratories involved in the SOHT proficiency testing program are forensic laboratories using hair analysis. Since 2001, Motherisk has exhibited (and continues to exhibit) very good performance for both its ELISA (pre-2010) and mass spectrometry based (post-2010) methods on proficiency testing, with results comparable to other laboratories around the world.154

9. In addition, the Hospital told the Independent Review that it was aware of and relied on MDTL’s participation and performance in the SoHT proficiency-testing scheme in considering the extent of oversight that was necessary for the Laboratory from 2006 to 2010. It also relied publicly on MDTL’s performance in the SoHT proficiency tests. Following the decision of the Court of Appeal in Broomfield, the Hospital’s president and chief executive officer, Dr. Michael Apkon, and its paediatrician-in-chief, Dr. Denis Daneman, wrote an opinion / commentary article in the Toronto Star, dated November 24, 2014, which relied on MDTL’s performance in the SoHT proficiency tests from 2001 to 2009, stating in part:

Since 2001, Motherisk has participated annually in international proficiency testing undertaken by an external body, to ensure that hair drug testing results are comparable to results generated in other laboratories around the world. The 2005 technique in question has produced no reported false-positive results with cocaine and no evidence of risk for over-estimation of cocaine concentrations in hair over the nine years of external proficiency testing conducted while it was in use. It is important to note that our methodology has evolved along with the science. Since 2010, we have used the most advanced methods available. At the time our methods changed, we used the “gold standard” (mass spectrometry) to validate the method used prior to 2010 (immunoassay). [Emphasis added.]

10. Given the importance that MDTL and the Hospital placed on the Laboratory’s proficiency test results, the Independent Review examined not only the results from 2001 to 2009, but also the forms that MDTL submitted to the SoHT, together with the underlying data (which were not submitted to the SoHT).

11. MDTL should not have represented its proficiency test results as evidence of the adequacy and reliability of its ELISA methodology. The reality is that, between 2001 and 2009, for most samples MDTL’s submissions to the SoHT did not reflect quantitative results that were actually generated by the Laboratory.155 First, in each year from 2001 to 2009, MDTL reported that its screening test results were obtained by ELISA, and its confirmation and quantification results were obtained by GC-MS.156 However, during this period, MDTL did not have the capability to, and did not, confirm its results in-house. Rather, MDTL sent its proficiency test samples to another laboratory for GC-MS or liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis.157 It did so even though such analysis was not part of the Laboratory’s standard procedure, and, in fact, it rarely sent any samples for confirmation testing. Moreover, MDTL’s submission forms to the SoHT did not disclose that its reported GC-MS results were not generated by MDTL, but by another laboratory.

12. Second, in the majority of its submissions to the SoHT (between 2001 and 2009), MDTL reported quantitative results that were not generated either in-house using ELISA or by its reference laboratory using GC-MS.158 In particular, for MDTL’s 2002–9 submissions to the SoHT, the Independent Review obtained the worksheets that Tatyana Karaskov, MDTL’s laboratory technician who carried out virtually all the Laboratory’s ELISA tests, had prepared using MDTL’s ELISA tests, as well as the results reports obtained from MDTL’s reference laboratory.159 These worksheets and results reports showed that, in some cases, MDTL appeared to report to the SoHT its ELISA test results for cocaine and benzoylecgonine (though it indicated on the proficiency test submission form that it had obtained those results using GC-MS); in other cases, MDTL reported its reference laboratory’s result (though sometimes rounded up or down to the nearest decimal); and still in others, the results that were reported to the SoHT as having been obtained using GC-MS analysis were not contained in either the ELISA worksheets or the reference laboratory’s results reports.

13. Table 7.1 identifies the quantitative results that MDTL reported to the SoHT from 2002 to 2009. Yellow indicates results that were found in MDTL’s ELISA worksheets (and presumably were MDTL’s ELISA results) but were reported as having been obtained by GC-MS; orange indicates results that were MDTL’s reference laboratory’s results; blue indicates results that MDTL had reported as having been obtained using GC-MS but were not contained in either the ELISA worksheets or the reference laboratory’s reports (and, as a result, the Independent Review could not determine where these results came from); purple indicates results that MDTL reported as ELISA results;160 and the white areas indicate where MDTL did not report a quantitative result (because the test was negative).

Table 7.1 Summary of quantitative results submitted by MDTL to the SoHT as having been obtained using GC-MS analysis, 2002–9a

Sample

Coc

BZE

Coca-
ethyl

Norcoc

6-AM

Morph

Cod

THC

Amph

Meth

MDMA

2002 Proficiency Test Submissions

A

2.65

0.80

1.03

0.54

0.24

B

3.70

0.56

0.42

0.21

0.40

1.56

0.30

0.50

C

2.60

1.96

0.20

Trace

Trace

D

0.55

E

2003 Proficiency Test Submissionsb

A

1.97

1.31

1.30

0.52

0.49

0.40

0.54

0.43

0.32

B

2.08

0.78

0.89

0.45

0.44

0.82

0.46

0.19

0.71

C

2.10

2.80

0.58

0.69

D

2004 Proficiency Test Submissions

A

B

2.95

0.62

0.67

0.40

0.55

1.23

4.96

C

37.53

12.66

1.52

D

E

2.61

0.89

3.60

1.35

2005 Proficiency Test Submissions

A

10.56

1.22

2.26

B

C

0.31

0.44

D

27.82

5.34

0.59

E

2.11

0.74

2.86

1.19

0.37

2006 Proficiency Test Submissions

A

0.78

0.41

7.8

1.9

0.4

B

1.27

0.55

0.168

1.03

C

11.19

7.92

D

E

2007 Proficiency Test Submissions

A

0.92

0.80

0.2

0.21

B

0.65

0.68

2.4

2.1

0.7

C

D

E

0.35

0.92

0.68

0.78

0.2

2008 Proficiency Test Submissions

A

7.97

3.86

4.92

3.50

1.54

0.14

B

C

7.36

2.97

0.8

0.3

0.22

0.5

1.4

2.2

D

4.80

2.07

4.42

2.30

1.54

E

1.51

0.5

0.15

2009 Proficiency Test Submissions

A

B

2.66

1.81

0.37

C

5.21

0.95

0.12

0.22

1.2

0.43

0.12

D

E

6.8

2.84

0.2

0.3

1.3

0.45

0.11

a The columns in this table represent results submitted for the following drugs or metabolites (in order): cocaine, benzoylecgonine, cocaethylene, norcocaine, 6-AM, morphine, codeine, cannabis / THC, amphetamine, methamphetamine, MDMA.

b In 2003, MDTL reported two results for cocaine, benzoylecgonine, and methamphetamines for Sample B (unwashed and washed) and two results for cocaine and benzoylecgonine for Sample C (unwashed and washed).

14. In short, and as is evident from Table 7.1, for the most part MDTL’s proficiency test submissions bore no resemblance to the actual test results generated by the Laboratory. There were only a handful of samples in a few years (yellow) in which MDTL reported its ELISA test result for cocaine and/or benzoylecgonine (though it reported to the SoHT that it had obtained those results using GC-MS). In the majority of cases (blue and orange), MDTL did not report results to the SoHT that were actually contained in its ELISA data. This practice was particularly true for the opiates (6-AM, morphine, and codeine).161 Simply put, for most samples, MDTL’s participation in the SoHT proficiency testing did not provide any assessment of the Laboratory’s testing methods.

15. Julia Klein, MDTL’s laboratory manager at the time, signed each of the proficiency test submission forms from 2001 until her departure in April 2005 (the 2005 proficiency test submission was not sent to the SoHT until after Ms. Klein’s departure). Ms. Klein told the Independent Review that, until at least 2004, she concluded from the discussions at SoHT meetings “that the purpose of the proficiency testing was more qualitative requiring a dichotomous positive / negative answer rather than quantitative requiring an exact concentration.” Although participation was not compulsory, Ms. Klein told the Independent Review that, for MDTL, “the purpose of participating was to verify if the methodology used in the lab was sufficiently accurate.” She described MDTL’s participation in SoHT proficiency testing in the following manner:

All ELISA testing was done by Tatyana Karaskov. She also couriered a portion of all the proficiency test samples to Psychemedics and in the more recent years to [United States Drug Testing Laboratory] for confirmation by GC-MS of some of the analytes. To the best of my recollection, whenever a GC-MS result was available, it was recorded on the proficiency test submission. When it was not available, the ELISA result was recorded on the submission. For example, if a proficiency sample contained amphetamine, methamphetamine, THC (cannabis), benzodiazepine which were not tested by GC-MS at that time, in those cases the ELISA result was recorded.162

16. Ms. Klein told the Independent Review that she instructed Ms. Karaskov to send the proficiency test samples to MDTL’s reference laboratory for GC-MS confirmation “to test for the individual analytes.” According to Ms. Klein, she gave this instruction to Ms. Karaskov “after deciding upon it with Dr. Koren.”163

17. Following Ms. Klein’s departure from MDTL in April 2005, Ms. Karaskov continued to prepare MDTL’s proficiency test submissions (from 2005 to 2009), but they were now signed by Dr. Koren. For his part, Dr. Koren denied knowing that MDTL was submitting results from another laboratory to the SoHT as part of MDTL’s proficiency test submissions. Dr. Koren told the Independent Review that, from 2005 to 2009, Ms. Karaskov completed the submissions and came to him to have him sign the cover sheet, which he did. He “was not aware of samples being sent to [MDTL’s reference laboratory] for GCMS testing” at the time, and he first learned of the practice in February 2015. According to Dr. Koren, Ms. Klein never discussed the practice of sending proficiency test samples to another laboratory with him, and he “never would have approved her practice if she asked.”

18. Ultimately, neither Dr. Koren nor Ms. Klein nor Ms. Karaskov was able to explain why MDTL was submitting results to the SoHT that bore no resemblance to the results generated by the Laboratory, and therefore could not be used to assess MDTL’s proficiency. Likewise, Dr. Koren, Ms. Klein, and Ms. Karaskov could not explain why MDTL sent proficiency test samples to a reference laboratory when that was not their standard procedure, or why a significant number of the results submitted to the SoHT did not appear in any of the underlying data. Dr. Koren acknowledged that MDTL’s practice of reporting the results obtained from its reference laboratory was “illogical and potentially misleading.”

19. Mr. Gareri did not have any involvement in MDTL’s proficiency test submissions until 2010, and he told the Independent Review that he discovered the reporting problems identified above only when he collected the 2001–9 submissions at the request of the Independent Review. Ultimately, between 2001 and 2009, MDTL’s proficiency test results did not attest to the Laboratory’s proficiency. Accordingly, they provide no assurance about the reliability of MDTL’s ELISA tests during those years.

2.2 Drugs of Abuse: Post-2010

20. Starting in 2010, Mr. Gareri began to sign the proficiency test submissions forms that were sent to the SoHT. By the time of the 2010 proficiency test submission, MDTL had transitioned to routine confirmation of its ELISA results. With one exception, it began to report its own GC-MS (and later LC-MS/MS) results to the SoHT. The one exception was in 2010, when MDTL reported that it had used GC-MS to obtain a result for cannabis / THC. At the time, MDTL did not have a GC-MS method for cannabis / THC, and it reported the cannabis / THC result that it had obtained from its reference laboratory (which turned out to be a false negative in one sample).

21. Mr. Gareri explained that MDTL reported its reference laboratory’s result for cannabis / THC in the Laboratory’s 2010 proficiency test submission because he was “evaluating the prospect of referring out all THC positive ELISA samples for GC-MS confirmation.” Mr. Gareri acknowledged to the Independent Review that he made an error in judgment by including another laboratory’s result in the submission, and he did not make the same mistake again after 2010.164 The Independent Review did not identify any of the earlier reporting problems in its review of MDTL’s 2011 to 2014 proficiency test submissions.

22. Accordingly, unlike the submissions from 2001 to 2009, MDTL’s proficiency test submissions from 2010 to 2014 (with the one exception in 2010 for cannabis / THC) actually measured MDTL’s performance as compared with the SoHT’s reference laboratories and other participating laboratories. Although MDTL’s qualitative results fared well (there were no false positives or false negatives),165 its quantitative performance in the 2010–14 SoHT proficiency tests was varied. By way of example, Table 7.2 provides a summary of MDTL’s quantitative performance on the proficiency test samples that were positive for cocaine and benzoylecgonine from 2010 to 2014.

23. As is evident from Table 7.2, there were material discrepancies between MDTL’s results and the median result of the SoHT’s reference laboratories, as well as the median result of all participating laboratories – in a material number of cases, the deviations were greater than 100 percent (in yellow) and, in one case, the deviation was in excess of three times (or 300 percent). In the end, MDTL’s performance in the SoHT profici­ency-testing scheme does not provide any assurance about the reliability of MDTL’s GC-MS procedure.

Table 7.2 Summary of MDTL’s performance (quantitative) in SoHT proficiency tests for cocaine and benzoylecgonine, 2010–14

Cocaine

Year

Samplea

MDTL Result

Reference Labs’ Medianb

Deviation from Reference Labs’ Median

Participating Labs’ Median

Deviation from Participating Labs’ Median

2010

A

6.87

5.4

27%

6.14

12%

2010

E

1.02

0.85

20%

1.02

0%

2011

Part I, Sample B

0.7

0.76

–8%

0.76

–8%

2011

Part I, Sample C

1.15

1.74

–34%

1.74

–34%

2011

Part II,
Sample A

4.72

4.9

–4%

5.36

–12%

2011

Part II,
Sample B

9.26

4.5

106%

4.55

104%

2012

Part I, Sample A

0.54

0.52

–4%

0.61

–11%

2012

Part I, Sample B

0.44

0.2

120%

0.2

120%

2012

Part I, Sample C

6.35

7.8

–19%

5.83

9%

2012

Part II, Sample B

5.38

5.9

–9%

5.54

–3%

2013

Part II,
Sample A

Trace

0.85

N/A

0.8

N/A

2013

Part II,
Sample C

3.14

4.2

–25%

4.2

–25%

2014

Part I, Sample A

1.18

1.3

–9%

1.47

–20%

2014

Part I, Sample B

6.24

3.1

101%

3.93

59%

2014

Part II,
Sample A

5.09

4.6

11%

4.22

21%

2014

Part II,
Sample B

2.36

2.1

12%

2.29

3%

Benzoylecgonine

Year

Samplea

MDTL Result

Reference Labs’ Medianb

Deviation from Reference Labs’ Median

Participating Labs’ Median

Deviation from Participating Labs’ Median

2010

A

9.42

3.9

142%

7.5

26%

2010

E

2.74

1.15

138%

1.48

85%

2011

Part I, Sample B

1.1

0.79

39%

0.78

41%

2011

Part I, Sample C

1.8

2.1

–14%

2.23

–19%

2011

Part II, Sample A

8.46

5.9

43%

6.02

41%

2011

Part II, Sample B

11.73

4.5

161%

4.07

188%

2012

Part I, Sample A

0.9

0.63

43%

0.81

11%

2012

Part I, Sample B

0.45

0.1

350%

0.17

164%

2012

Part I, Sample C

14.44

10.2

42%

11.37

27%

2012

Part II, Sample B

8.03

6.2

30%

6.89

17%

2013

Part II, Sample A

0.87

1.8

–52%

1.72

–48%

2013

Part II, Sample C

5.46

6.6

–17%

6.26

–13%

2014

Part I, Sample A

3.52

3.5

0.6%

3.56

–1%

2014

Part I, Sample B

6.77

4.0

69%

4.72

43%

2014

Part II, Sample A

6.37

5.8

10%

6.12

4%

2014

Part II, Sample B

3.48

3.5

–0.6%

3.22

8%

a Starting in 2011, the SoHT began to offer proficiency testing twice a year, in two parts.

b The reference laboratories’ median was estimated from the results charts.

2.3 Alcohol Markers

24. The SoHT has offered proficiency testing for FAEEs for only four years – from 2011 to 2015166 – and MDTL participated in the program from 2011 to 2014.167 During MDTL’s participation, the Laboratory’s results deviated significantly from the median results of the participating laboratories. That said, MDTL was not unique in this regard – there appeared to be wide variation across participating laboratories’ results in 2011, as well as in subsequent years.168 This wide deviation may explain why some hair-testing experts have since moved away from FAEEs analysis. In the end, however, as with MDTL’s proficiency test results from 2010 to 2014 for drugs of abuse, its performance on the FAEEs proficiency tests does not provide any further assurance about the reliability of its FAEEs method.

3. MDTL and Accreditation

25. In Ontario, accreditation (including participation in proficiency testing) is mandatory for clinical laboratories but not a requirement for laboratories providing forensic services. If a laboratory can otherwise demonstrate that it has adhered to internationally recognized forensic standards in its processes and procedures, the fact that it is not accredited is not determinative of whether its test results may be admissible in court or are otherwise reliable for forensic purposes.

26. However, since 2006, accreditation under a “quality management program” is statutorily mandated for all laboratories that provide clinical tests (i.e., tests used for diagnostic, prophylaxis, or treatment purposes).169 This mandatory accreditation may be obtained under the Ontario Laboratory Accreditation program, which is administered by the Institute for Quality Management in Healthcare (IQMH), and formerly administered by the Ontario Medical Association.

27. In January 2011, it participated for the first time with SickKids’s clinical laboratories in the Ontario Laboratory Accreditation program. MDTL relied on its accreditation under this program to bolster the reliability of its tests; however, for the reasons set out below, the fact that MDTL passed the accreditation inspection under this program does not give me any further assurance about the adequacy and reliability of its analytical tests or interpretations.

3.1 No Forensic Accreditation

28. There are bodies in Canada and the United States that accredit laboratories to forensic standards. For example, the Standards Council of Canada, which is a federal Crown corporation, provides accreditation for forensic laboratories, including at some sites of the Royal Canadian Mounted Police Forensic Science and Identification Services, and the Laboratoire de science judiciaires et de médecine légale in Montreal. In the United States, one of the accrediting bodies for forensic laboratories is the American Society of Crime Laboratory Directors / Laboratory Accreditation Board (ASCLD / LAB), which is a not-for-profit corporation specializing in the accreditation of public and private crime laboratories. The ASCLD / LAB accredits, among many other laboratories, the central laboratory of the Centre of Forensic Sciences in Toronto and the Northern Regional Laboratory in Sault Ste. Marie, Ontario.

29. Organizations that offer forensic accreditation, such as the Standards Council of Canada and the ASCLD / LAB, typically accredit laboratories based on a standard created by the International Standards Organization (ISO),170 ISO 17025:2005, General requirements for the competence of testing and calibration laboratories, as well as additional requirements that are targeted to the specific needs of forensic laboratories. ISO 17025 is different from, and contains more stringent requirements than, the standards that are applied in the accreditation of clinical laboratories.171

30. MDTL has never been accredited as a forensic laboratory. Although it is not necessarily unusual for a hair-testing laboratory to lack forensic accreditation,172 and although forensic accreditation is not required for a laboratory to carry out tests for forensic purposes in Ontario, MDTL’s lack of forensic accreditation meant that no accreditation body ever assessed the work of the Laboratory from a forensic perspective.

3.2 Accreditation under the Ontario Laboratory Accreditation Program

31. In January 2011, MDTL participated for the first time in an accreditation inspection under the Ontario Laboratory Accreditation program, which offers accreditation to laboratories providing clinical tests in the province. To understand what accreditation means under this program, it is important first to understand the clinical laboratory licensing framework in the province, including how accreditation fits into that framework.

3.2.1 Licensing and Accreditation Framework

32. Since 1972, in order to operate as a clinical laboratory in Ontario, a laboratory must be licensed.173 Throughout the period under review, the Laboratory and Specimen Collection Centre Licensing Act (Laboratory Licensing Act) required that a laboratory performing clinical tests be licensed by the director of laboratory and specimen collection centre licensing, who is appointed by the Minister of Health and Long-Term Care.174

33. In 2000, the Ministry of Health and Long-Term Care (the Ministry) contracted with the Ontario Medical Association (OMA) to create and implement a mandatory medical laboratory accreditation program to assess and rate licensed laboratories in accordance with established criteria. The OMA began phasing in the accreditation program in 2003. Since that time, the accrediting body, now called the Institute for Quality Management in Healthcare (IQMH), has accredited laboratories offering clinical testing in the province. For ease of reference, I refer to both the IQMH and its predecessor as “IQMH.”

34. In 2006, the Laboratory Licensing Act was amended to require, as a condition of licensing, that a laboratory “meet the requirements of a quality management program” – a condition that is satisfied by participation in the Ontario Laboratory Accreditation program. Accordingly, since 2006, accreditation under a “quality management program” has formed an express condition of a laboratory’s licensing.175

35. Laboratory licences are issued to a facility, not to individual laboratories within a facility. Accordingly, although a hospital such as SickKids may have several clinical laboratories, it will have only one laboratory licence that covers all its clinical laboratories. Under the licensing scheme of the Ministry, each facility is licensed to perform certain tests, or classes of tests. The tests that a facility is licensed to perform are identified on the licence in the form of test codes.

36. Under the Ontario Laboratory Accreditation program, the IQMH established a series of more than 500 requirements that must be met in order for a clinical laboratory to demonstrate that it complies with the conditions for accreditation and therefore licensing in Ontario. These requirements, which are based on three sets of standards developed by the ISO for medical laboratories,176 fall into 10 categories: (1) Organizational Structure, Personnel Policies and Training, and Laboratory Management; (2) Quality Management System; (3) Physical Facilities; (4) Equipment, Reagents and Supplies; (5) Pre-analytical Process; (6) Analytical Process; (7) Quality Assurance; (8) Post-Analytical Process (Reporting); (9) Laboratory Information System; and (10) Safety.

37. Compliance with these more than 500 requirements is assessed at a site visit (also called an inspection) by a team of assessors, which includes physicians, scientists, and medical laboratory technologists. An assessor who finds a non-conformance with one of the requirements must determine whether it is “minor” (an isolated incident) or “major” (a systemic non-conformance, or an isolated incident that has an impact on patient safety or laboratory results). Laboratories have 90 days to address the findings of an assessment inspection. They are required to correct any major non-conformances fully and submit an action plan for the correction of any minor non-conformance.

38. Although laboratory licences must be renewed annually (and, on renewal, the laboratory must inform the Ministry of the volume of tests it has conducted under each test code over the past year and advise the Ministry of any tests to be added to, or deleted from, the licence), the inspection under the Ontario Laboratory Accreditation program is required only once every four years.177 However, laboratories are required to submit a self-assessment at the midterm mark, in which the laboratory assesses its own procedures and tests according to the program’s requirements and submits the self-assessment report to the IQMH.178

39. All laboratories performing tests for clinical purposes must be included in the accreditation assessment. If accreditation is granted, the facility receives an accreditation certificate along with a scope of accreditation that lists the tests for which the facility is accredited. (These tests are categorized by discipline – for example, biochemistry, immunology, virology – and identified by test code. The facility is then licensed to carry out tests under these test codes.) In a facility with multiple laboratories (such as SickKids), the scope of accreditation does not identify which tests are licensed for each laboratory within the facility.

3.2.2 Accreditation of MDTL under the Ontario Laboratory Accreditation Program

40. In 2006, the clinical laboratories at SickKids participated for the first time in the Ontario Laboratory Accreditation program. However, MDTL was not included in that assessment, likely because it was considered by the Hospital to be a research laboratory.

41. In 2008, the Hospital’s clinical laboratories participated in the midterm self-assessment that was required under the Ontario Laboratory Accreditation program. Again, MDTL was excluded from that self-assessment.

42. It was not until January 2011, when SickKids was next assessed by the IQMH, that MDTL first participated in the Ontario Laboratory Accreditation assessment. The Hospital’s accreditation assessment visit took four days, with the IQMH assessors spending close to two full days with MDTL staff, including Mr. Gareri, quality manager Netta Fulga, and laboratory supervisor Dr. Bhushan Kapur.

43. At the conclusion of the January 2011 assessment visit, the IQMH assessment team identified three minor non-conformances at MDTL, which the Laboratory dealt with promptly, and the Hospital’s certificate of accreditation (which included MDTL) was issued on June 9, 2011. A midterm self-assessment was completed by the Hospital and submitted to IQMH in 2013, and, in December 2014, MDTL was inspected as part of the next accreditation cycle for the Hospital. The 2014 accreditation assessment did not identify any non-conformances by MDTL.

44. Although MDTL passed the Ontario Laboratory Accreditation assessment in January 2011 and again in December 2014, its accreditation under the Ontario Laboratory Accreditation program does not provide any assurance about the adequacy and reliability of MDTL’s analytical methods and interpretations for use in child protection and criminal cases throughout the period of the Independent Review for three reasons.

45. First, the Ontario Laboratory Accreditation was, by definition, a clinical laboratory accreditation. As noted above, no external body has ever evaluated MDTL’s compliance with forensic standards.

46. Second, MDTL’s first accreditation assessment took place in January 2011, after MDTL had transitioned to routine GC-MS confirmation testing for most drugs and metabolites for adults and children (in September 2010) and fully implemented its quality management system (in December 2010). Accordingly, the fact that MDTL passed its assessment in January 2011 does not speak at all to its methods and procedures before that date. In fact, given the serious flaws in MDTL’s practices before January 2011, MDTL would not have passed its accreditation assessment even if it had participated in the process sooner (for example, in 2006 for the Hospital’s first assessment or in 2008 for the Hospital’s midterm self-assessment).

47. In these circumstances, MDTL should not have described its accreditation to the court in the manner it did on the Broomfield appeal. In his January 11, 2013, report, which was submitted on the Broomfield appeal, Dr. Koren stated:

MDTL is accredited by OLA (Ontario Laboratory Accreditation). OLA is part of the Standards Council of Canada, and OLA standards are based on the International Organization for Standardization (ISO) criteria. We passed the accreditation with outstanding marks, being the first drug hair testing laboratory to do so in Canada. The tests we conducted as part of the accreditation further validated our ELISA cocaine and BZE tests as accurate, sensitive and specific. [Emphasis added.]

48. This description of the accreditation that MDTL obtained overstates the relationship between accreditation and the reliability of MDTL’s ELISA test results. MDTL was not included in the Hospital’s clinical accreditation, and had not participated in any accreditation assessment, at the time of the hair testing in Broomfield. Nor is it accurate to suggest that the accreditation process “validated [the] ELISA cocaine and BZE tests as accurate, sensitive and specific.” When the January 2011 assessment took place, MDTL was no longer using ELISA to quantify cocaine and benzoylecgonine concentrations in hair – it had moved to GC-MS. Contrary to Dr. Koren’s suggestion, no accrediting body has ever assessed MDTL’s ELISA tests for quantitative purposes.

49. Third, even at the January 2011 assessment, the IQMH did not assess the robustness of MDTL’s analytical methods or interpretations. In particular, the IQMH described its accreditation assessment to the Independent Review as “horizontal and holistic,” rather than as a “vertical audit” of every accreditation requirement against each of the laboratory’s procedures. This approach means, for example, that assessors do not examine every standard operating procedure that a laboratory has adopted, nor do they read every line of the procedures they do examine. When reviewing a laboratory’s analytical methods, an assessor will generally select a subset of the laboratory’s standard operating procedures and review those procedures to ensure they are complete and that the method adopted meets the Ontario Laboratory Accreditation requirements. Those requirements mandate that the test (1) is based on a method cited in a published peer-reviewed text or journal; (2) is recommended by international, national, or regional guidelines; or (3) if it is a method developed in-house, has been properly validated for its intended use.179

50. MDTL’s methods fell under the third category; accordingly, the Laboratory was required to demonstrate that its in-house methods had been properly validated for their intended use by showing that it had validation documents (which it did). However, the assessment team was not charged with evaluating, and it did not evaluate, the validity or reliability of the Laboratory’s analytical methods or results.180

51. Rather, the IQMH assessor who was responsible for MDTL’s accreditation visit in January 2011 explained to the Independent Review that the assessment team looks to ensure that a laboratory has a complete quality management system as well as standard operating procedures that are being followed. Accordingly, by way of example, the assessment team would not have considered whether 65 percent was a low-match for identifying drugs in hair when operating the GC-MS instrument in full scan mode, nor would it have assessed the reliability of using such a match.181 Rather, the assessment team checked only to ensure that MDTL was in fact applying its own 65 percent match criteria. Similarly, the team would not have assessed the adequacy of MDTL’s reliance on ELISA for qualitative purposes for the five drugs for which the Laboratory had not developed a GC-MS method, and for all neonate samples.182 In other words, the team was not assessing the adequacy and reliability of the Laboratory’s procedures but looking only to see if MDTL had analytical procedures and if it was complying with them.

52. Nor did the IQMH assessment team consider the manner in which the Laboratory was interpreting results for its customers. The assessor in charge of MDTL’s assessment in January 2011 told the Independent Review, for example, that he did not review the concentration ranges that MDTL used to interpret its tests results for drugs of abuse. Accordingly, the Ontario Laboratory Accreditation assessment did not assess the reliability and adequacy of either MDTL’s analytical methods or its interpretations at any time during the period of the Independent Review.

53. In short, although MDTL relied on its participation in the SoHT’s proficiency-testing program as well as its incorporation into the Hospital’s clinical accreditation framework to support the robustness of its hair tests, it should not have done so.


Notes

152 Starting in 2011, the SoHT began to offer its proficiency tests twice a year.

153 Some laboratories may offer only screening tests, in which case they will submit only the qualitative result from the screening test.

154 OACAS information sheet for distribution to all members, December 3, 2014.

155 Table 7.1 identifies all the quantitative results that MDTL reported to the SoHT from 2002 to 2009 and where those results originated.

156 In some years, MDTL reported to the SoHT that it was using ELISA to quantify amphetamines, methamphetamines, and cannabis / THC.

157 It is an accepted and indeed common practice for a laboratory to send samples to another laboratory (known as a reference laboratory) for testing. Such referral typically occurs when a laboratory does not have the equipment or the expertise to carry out the test that has been requested. It may also occur as a way of double-checking the laboratory’s own results. The reference laboratory in this context is subcontracted to carry out the test. However, the SoHT’s instructions were to analyze the proficiency test samples following the standard procedures used routinely in the Laboratory.

158 As noted, the SoHT proficiency test submission forms required participating laboratories to submit both qualitative results (obtained from a screening test) and quantitative results (obtained from a confirmation test). MDTL reported the qualitative results generated from its ELISA test (which was a screening test), but it did not always report the ELISA test’s quantitative results.

159 MDTL was unable to locate any of the underlying worksheets or results reports for its 2001 submission.

160 The results that MDTL reported as having been obtained using ELISA were, for the most part, contained in the ELISA worksheets that MDTL produced.

161 It is not surprising that MDTL would not report its ELISA results for the opiates because the ELISA kit that MDTL used could not distinguish among 6-AM, morphine, and codeine.

162 Psychemedics and the United States Drug Testing Laboratory were the reference laboratories that MDTL used during this period.

163 Ms. Karaskov also confirmed that she recorded the quantitative results from MDTL’s reference laboratory on MDTL’s proficiency-testing submissions.

164 It was not necessarily incorrect to submit the results of another laboratory if MDTL was routinely sending its samples for cannabis / THC analysis. The error in judgment, therefore, is not the submission of another laboratory’s results, per se, but rather the failure to follow MDTL’s standard procedures in its proficiency test submissions. Starting in 2011, MDTL reported both its ELISA screening test (qualitatively) and GC-MS confirmation test results (quantitatively) on its proficiency test submissions. For cannabis / THC, for which MDTL did not have GC-MS capability, it reported the quantitative result as “NA.”

165 The one exception was the false negative for cannabis / THC in 2010, which turned out to be MDTL’s reference laboratory’s result.

166 The SoHT postponed the FAEEs proficiency-testing program indefinitely in June 2015.

167 In 2010, prior to the establishment of the SoHT’s FAEEs proficiency-testing program, MDTL participated in an inter-laboratory comparison with Professor Fritz Pragst’s laboratory at the Institute of Legal Medicine and Forensic Sciences, University Hospital Charité, in Berlin, and Dr. Silke Suesse’s laboratory at Trimega Laboratories in Ulm, Germany.

168 The variation meant that some laboratories would have interpreted the results from a particular sample as evidence of social drinking or abstinence, while another laboratory may have reported the sample results as being consistent with chronic and excessive alcohol consumption.

169 Laboratory and Specimen Collection Centre Licensing Act, RSO 1990, c L.1, s 9(14) (Laboratory Licensing Act).

170 The ISO is an independent, non-governmental member-based organization comprised of 163 national standards bodies. The ISO is the world’s largest developer of voluntary international standards and has published more than 19,500 international standards on a variety of subject areas, including standards for laboratories. The ISO does not accredit laboratories. It promulgates the standards that accrediting bodies use to evaluate laboratories.

171 Though not specific to forensic laboratories, ISO 17025 is recognized internationally as the appropriate standard for laboratories performing forensic testing. It contains requirements for testing laboratories to demonstrate their ability to generate valid test results. It consists of 11 tenets and also specific requirements that detail how the tenets are met, including the assessment of laboratory documentation, the validity and appropriateness of test methods, proficiency testing, measurement traceability, the technical competence of staff, the testing environment, and measurement uncertainty calculations.

172 A survey conducted by the SoHT in 2007 revealed that, of the 52 hair-testing laboratories that responded to the survey, 27 were not forensically accredited: G. Cooper, M. Moeller, and R. Kronstand, “Current Status of Accreditation for Drug Testing in Hair” (2008) 176 Forensic Science International 9–12.

173 Public Health Act, RSO 1970, c 377, as amended by The Public Health Amendment Act, 1972, SO 1872, c 80.

174 Laboratory Licensing Act, s 6, 9(1). The Laboratory Licensing Act defines “laboratory” as “(a) an institution, building or place in which operations and procedures for the microbiological, serological, chemical, hematological, biophysical, immunohematological, cytological, pathological, cytogenetic, molecular genetic or genetic examination, or such other examinations as are prescribed by the regulations, of specimens taken from the human body are performed to obtain information for diagnosis, prophylaxis or treatment, and (b) any other institution, building or place that may be provided for in the regulations”; Laboratory Licensing Act, s 5.

175 Prior to the 2006 amendment, the Laboratory Licensing Act required, as a condition of licensing, that “the performance of tests in the laboratory meet the generally accepted standards of proficiency in such tests”: s 9(14), RSO 1990, c L.1.

176 Although the ISO refers to “medical laboratories,” I do not consider there to be a material distinction between medical laboratories and clinical laboratories for the purposes of this Report. The three ISO standards on which the Ontario Laboratory Accreditation requirements are based are ISO 15189:2012, Medical laboratories – requirements for quality and competence; ISO 15190:2003, Medical laboratories – requirements for safety; and ISO 22870:2006, Point-of-care testing – requirements for quality and competence.

177 From 2003, when the Ontario Laboratory Accreditation program was rolled out, until 2008, it was once every five years.

178 In addition to the accreditation mandated by the Laboratory Licensing Act, since 2012 IQMH has offered voluntary accreditation under its ISO 15189 Plus program. The ISO 15189 Plus accreditation requires an additional mid-cycle assessment visit (instead of the midterm self-assessment under the Ontario Laboratory Accreditation program).

179 Requirement VI.1, Institute for Quality Management in Healthcare (IQMH), Accreditation requirements, version 6.0, December 2013; Requirement VI.1. Ontario Laboratory Accreditation (OLA) Requirements, version 4.1 (released July 2008).

180 According to the IQMH assessor, the reliability of a laboratory’s results or outcomes is assessed through peer review in scientific journals (if the method has been published) and/or by participation in proficiency testing. (Proficiency testing is a requirement of accreditation under the Ontario Laboratory Accreditation program.)

181 As discussed in Chapter 5, the accepted practice according to the Independent Review’s experts is to require a match of 95 percent or higher when the GC-MS instrument is run in full scan mode.

182 As discussed in Chapter 5, the five drugs for which MDTL did not develop a GC-MS method were benzodiazepines, barbiturates, PCP, LSD, and cannabis / THC.

Chapter 8
What Went Wrong at SickKids

1. Introduction

1. In the preceding chapters, I concluded that the analytical methods and interpretations of the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) were inadequate, unreliable, and did not meet internationally recognized forensic standards during the period under review. At the same time, participants in the justice system – including child protection workers, police officers, lawyers, and the courts – expected and relied on MDTL to provide adequate and reliable hair tests and opinions for use in child protection and criminal cases. MDTL did not live up to those expectations.

2. MDTL was a laboratory housed in one of the most prominent and well-respected children’s hospitals in Canada – the Hospital for Sick Children (SickKids or the Hospital). In this chapter, I consider MDTL’s deficiencies on an institutional level, asking the question: What went wrong at SickKids?

2. Failure to Appreciate the Forensic Nature of MDTL’s Hair Tests

3. Throughout the 10-year period under review, MDTL was testing an average of 2,000 hair samples each year for child protection agencies, and from time to time its hair tests were also used in criminal proceedings. The hair tests were forensic in nature, and the service that MDTL was offering to customers was a forensic one. However, neither MDTL nor the Hospital appears to have appreciated that the Laboratory was engaged in forensic work and that it was required to meet forensic standards.

2.1 The Views of MDTL’s Leadership

4. MDTL’s laboratory director and laboratory managers told the Independent Review they did not consider the Laboratory to be performing forensic work at any time during the period under review. Instead, they insisted that MDTL’s work was purely of a clinical nature. Dr. Gideon Koren, MDTL’s laboratory director, said that he considered MDTL to be providing “clinical information” to child protection agencies, which was being used for “treatment purposes and ensuring child safety.” According to him, the fact that only a small minority of cases went to court meant that “the vast majority served as a treatment tool of therapeutic drug monitoring.”

5. In support of his view that MDTL was performing clinical, not forensic, toxicology work, Dr. Koren advised the Independent Review that the term “forensic” was not mentioned by any judge, child protection lawyer, defence lawyer, or Crown lawyer and that the Hospital’s focus was on clinical (rather than forensic) accreditation:

It is noteworthy that [the Hospital] asked that MDTL undertake accreditation as a clinical laboratory. The term forensic or any expectation for forensic standards were not mentioned once by any leader, in [the Department of Paediatric Laboratory Medicine] or the rest of [the Hospital], including risk management or the executive team. All of them were well aware of our activities with [children’s aid societies] and the court. [Emphasis in original.]

6. Joey Gareri, MDTL’s laboratory manager from May 2005 onward, also reported that he did not consider the hair testing that MDTL carried out for child protection agencies to be forensic work. He told the Independent Review that he considered “forensic” to refer to criminal cases, which made up only a tiny percentage of MDTL’s work. With respect to the hair tests that were done at the request of child protection workers, Mr. Gareri said that those tests formed part of the “case management” process adopted in child protection cases – for example, to provide a child protection agency with a means of assessing whether a parent was telling the truth about past or current drug use as part of the agency’s ongoing dialogue with the parent. For her part, Julia Klein, MDTL’s laboratory manager until April 2005, told the Independent Review that MDTL was not a forensic laboratory and that its work was “definitely not forensic.”

7. The only member of MDTL’s management who acknowledged that MDTL was engaged in forensic work was Dr. Bhushan Kapur, MDTL’s laboratory supervisor. He was appointed to the laboratory supervisor position only in October 2009, however, and his scope of responsibility was limited primarily to method development and quality control. He was not involved in carrying out the hair tests, reporting the results, or interpreting them for customers or other users of the tests, including child protection workers, police officers, lawyers, or the courts.

8. Whether a hair test is clinical or forensic does not turn on whether the test result is ever tendered as evidence in a court proceeding, criminal or otherwise. A hair test can be forensic even where no court proceeding is ever initiated. What distinguishes a clinical test from a forensic test is the purpose behind the test: If the test is carried out or is used for a legal purpose, then it is a forensic test. I do not accept Dr. Koren’s and Mr. Gareri’s descriptions of the nature of MDTL’s hair analysis as having a “treatment,” “therapeutic drug monitoring,” or “case management” purpose. The fact that a hair test is carried out at the request of a child protection agency for the purposes of a potential or ongoing child protection proceeding – which can have legal implications for the individuals involved – takes the test outside the clinical, and into the forensic, realm. Moreover, the reasoning that others in the justice system or within the Hospital reportedly did not use the term “forensic” does not detract from this conclusion.

9. Given the manner in which MDTL operated, Dr. Koren, Mr. Gareri, and Ms. Klein should have appreciated that MDTL was carrying out forensic, not clinical, work. Although some hair tests may have been requested by physicians for clinical purposes, MDTL was carrying out tests on an average of 2,000 hair samples each year at the request of Ontario child protection agencies for use in potential or existing child protection cases (see Appendix 7). Requests from child protection agencies were the bread and butter of MDTL’s hair-testing service; they were not one-offs.

10. Indeed, the failure of MDTL to appreciate that it was doing forensic work is exacerbated by the fact that it actively marketed its services to child protection agencies:

  • Mr. Gareri frequently attended at child protection agencies to provide education and training to child protection workers on the use of MDTL’s hair test results in the child protection context.183
  • Starting in December 2005, MDTL published its first annual “Motherisk Drug Testing Newsletter for Children’s Aid Societies,” which, in its own words, was “developed especially for Canada’s children’s aid societies.” MDTL published additional newsletters dedicated to child protection agencies in December 2006, December 2007, and December 2009.184
  • Beginning in at least August 2006, MDTL offered preferential rates (which it called “social services” rates) to child protection agencies for the hair tests it carried out for drugs of abuse.185
  • MDTL staff routinely received calls from child protection workers requesting interpretations of the Laboratory’s hair test results. In addition, Mr. Gareri was regularly called to testify in court about MDTL’s test results.

11. Although Dr. Koren, Mr. Gareri, and Ms. Klein denied knowing that MDTL was carrying out forensic tests, they nonetheless developed some laboratory practices for hair tests that were clearly not clinical in nature. For example, throughout the period under review, MDTL used a Chain of Custody Requisition Form for samples that were collected outside the Hospital as a way to maintain chain of custody of the sample. In addition, sample collectors were asked to seal the envelope containing the hair sample and to sign the seal before sending it to the Laboratory for analysis. These chain-of-custody procedures have no place in a clinical laboratory but are a necessary feature of a forensic one. By using chain-of-custody forms, MDTL created the illusion that it was a forensic laboratory, employing standard forensic procedures. As a result, not only did MDTL fail to provide any caveat to participants in the justice system that its results were not generated in a forensic setting but some of its practices actually suggested the opposite.

12. Mr. Gareri explained MDTL’s use of the Chain of Custody Requisition forms by saying that, although MDTL did not consider itself to be a forensic laboratory, it still sought to meet forensic standards where possible. Mr. Gareri also told the Independent Review that he had aspirations to seek forensic accreditation down the line (even though no one had asked MDTL to become forensically accredited). Ultimately, the fact that MDTL’s leadership turned its mind to the use of forensic standards suggests that it was aware to some degree that its work was forensic in nature.

13. Moreover, MDTL acknowledged that some of its tests were for “forensic purposes” in at least one of its internal documents. MDTL’s 2006/2007 Business Plan noted that MDTL was a “one of a kind centre” in part because it carried out tests “for forensic purposes”:

One of a Kind Centre – Motherisk Drug Testing Lab is the only one of its kind in Canada. Only one of very few in the world dealing with pediatric hair samples. There are in total only 20 labs in the world that perform these types of tests, but most of the labs are forensic labs. Motherisk Drug Testing Lab not only performs tests for forensic purposes but is also open to the public, making this lab unique and accessible to a greater clientele. [Emphasis added.]

14. In a booklet MDTL published in 2012 entitled The Motherisk Guide to Substance Abuse Monitoring, MDTL also described itself as translating “established forensic toxicology methods” to children and neonates:

Over the years the Motherisk Lab has translated established forensic toxicology methods (e.g. hair analysis) and applied them to children and neonates, as well as validating novel neonatal tests (e.g., meconium analysis) and advancing the use of alcohol biomarkers in adult hair to establish drinking history.

15. These documents suggest that there was some recognition on the part of MDTL’s leadership that the Laboratory’s hair analysis developed from “forensic toxicology methods” and that the hair tests it carried out for use in child protection and/or criminal cases had or could have “forensic purposes.” However, notwithstanding that recognition, the Laboratory’s leadership failed to adequately identify the forensic nature of the Laboratory’s work and the need for the Laboratory to comply with forensic standards. The result was inevitable: MDTL’s operations fell woefully short of internationally recognized forensic standards.

2.2 The Views of the Hospital’s Leadership

16. MDTL was originally a research laboratory, housed in the SickKids Research Institute. The Hospital told the Independent Review that it was not until 2005 that it became aware of the nature and quantity of the testing that the Laboratory was carrying out for non-research purposes. When the Hospital’s leadership became aware that MDTL was routinely carrying out hair tests for child protection cases, however, the Hospital viewed MDTL’s work as being a shift from research toward clinical practice. There is no evidence that anyone in a leadership position at the Hospital gave any meaningful consideration to the reality that MDTL was, for the most part, carrying out hair tests for forensic, not clinical, purposes. Nor was meaningful consideration given to the need to ensure that MDTL was complying with forensic standards or perhaps the need to stop MDTL from continuing to test samples for forensic purposes.

17. To the contrary, as described below, notwithstanding its knowledge about the nature and quantity of MDTL’s hair tests, SickKids focused on bringing MDTL into its clinical licensing and accreditation process, and it did so slowly and in the absence of any meaningful oversight of the Laboratory’s work.

3. Inadequate Training and Expertise at MDTL

18. None of the Laboratory’s leadership had any formal training in forensic toxicology. Dr. Koren, who was the director of the Laboratory, is a physician and clinical toxicologist, with no training or any experience in a forensic toxicology laboratory.

19. In addition, the laboratory manager during the bulk of the period under review, Mr. Gareri, was a master of science student when he was first appointed to the position in May 2005. Not only did he lack forensic training but he had no experience at all with hair testing. Although Mr. Gareri made efforts to learn on the job,186 it is apparent that, at least in his first few years, he did not have the training and expertise to understand either the analytical methods that MDTL used at the time or their limitations. Mr. Gareri’s focus when he first started as laboratory manager was on keeping the Laboratory running after Ms. Klein’s departure and on overhauling its pre- and post-analytical procedures, which were eventually implemented in July 2007.187 Mr. Gareri told the Independent Review that he “did not get into the analytics of what the lab was doing” during that time; instead, he deferred to Dr. Koren and Tatyana Karaskov, MDTL’s laboratory technician who carried out virtually all MDTL’s hair tests using enzyme-linked immunosorbent assay (ELISA) tests, on matters relating to hair analysis.

20. Notwithstanding his lack of training and expertise in hair analysis and his lack of involvement in the analytical methods employed by the Laboratory, Mr. Gareri was given primary responsibility for providing interpretation opinions to customers – the majority of whom by far were child protection workers – almost immediately after he began working at the Laboratory.188 Beginning sometime in 2006, Mr. Gareri also began to testify in court about MDTL’s hair test results, mostly in child protection cases.189

21. However, opinions interpreting hair test results are expert opinions. Given the many complexities in hair analysis, they must be provided by individuals who have the appropriate training and expertise to appreciate the nuances and limitations in the analysis as well as the controversies in the science. Mr. Gareri lacked the expertise required to fulfill that role in three material respects:

  • Although the person who interprets the hair test result need not be the one who carries out the analytical test, she or he must have sufficient knowledge of the analytical principles underlying that test to provide an expert opinion about the result. Mr. Gareri lacked this knowledge when he became MDTL’s laboratory manager.
  • Mr. Gareri lacked training and experience on the interpretation of MDTL’s hair test results, which can be complex and can depend on a variety of factors, both inside and outside the laboratory’s control.
  • Mr. Gareri lacked the training and experience needed to understand the role of an expert witness, including the need to be independent and objective and to explain any controversies or limitations in the science to users of the test results. Indeed, Mr. Gareri told the Independent Review it was not until recently, in the context of this Review, that he learned about the need for an expert to lay out the controversies in the scientific field for the end user, whether it is a child protection worker or the court. This revelation means that he provided expert opinions to customers and other users of MDTL’s hair test results for almost a decade under a misapprehension about the proper role of an expert witness.

22. From his perspective, Mr. Gareri acknowledged to the Independent Review that he should not have been hired as laboratory manager in 2005:

Ultimately, [I] had no training in management and only [my] Master’s degree research experience in analytics when [I] took over management of the laboratory in 2005. The reality is that [I] learned “on the job” on a job that you should not be learning on. In retrospect, [I] would not have hired [my]self back in 2005 to act as the Laboratory Manager unless a formal training program was in place (i.e. taking over after being trained by an existing manager for an extended period of time).

23. I agree with Mr. Gareri’s comments about the position in which he was placed. Ultimately, MDTL’s leadership appointed Mr. Gareri to a job he was not qualified to fill, gave him primary responsibility for giving expert opinions on matters in which he lacked sufficient expertise, and failed to provide him with the training needed for the role.

24. In addition, MDTL’s use of graduate students and fellows (whom MDTL called “laboratory counsellors”) to provide interpretations over the telephone to customers also demonstrated a fundamental misunderstanding of the nature of the opinions that were being provided. None of those laboratory counsellors was qualified to provide an expert opinion on the interpretation of hair test results at any time during the period under review.

4. Slow Transition from Research to Clinical Laboratory, but Never to Forensic Laboratory

25. Even putting aside SickKids’s failure to appreciate the forensic nature of MDTL’s hair tests, the Hospital took an inordinately long time to bring MDTL into compliance with the standards required to ensure the adequacy and reliability of its results for clinical purposes.

26. Once it learned of the extent of the tests that MDTL was performing for non-­research purposes in 2005, the Hospital considered the need to bring MDTL into SickKids’s clinical licensing and accreditation process. Nevertheless, MDTL was not included in the Ontario Laboratory Accreditation inspection that the Hospital’s clinical laboratories underwent in 2006. Nor was it included in the midterm self-assessment that the Hospital’s clinical laboratories underwent in 2008.

27. In fact, in 2008, when the Department of Paediatric Laboratory Medicine (DPLM) was performing its midterm self-assessment, it specifically turned its mind to MDTL’s inclusion in that process, and MDTL prepared some documents for DPLM for that purpose. Ultimately, however, the Hospital decided not to include MDTL in the midterm self-assessment. According to the Hospital, “while initially DPLM and MDTL aimed to bring MDTL into Hospital [Ontario Laboratory Accreditation] accreditation preparation as part of the 2008 mid-term self-assessment, it was recognized that this may not be achieved, and sights were set on the December, 2010 accreditation inspection instead.”190

28. It is not surprising that the Hospital recognized that clinical accreditation for MDTL might not be achieved in 2008. If MDTL had been included in the midterm self-­assessment at the time, it would have been clear to the accrediting body that MDTL did not meet the requirements of the Ontario Laboratory Accreditation program for a clinical laboratory. The stark reality is that, until 2010, MDTL did not have a quality management system or any standard operating procedures for its hair tests for drugs of abuse, both of which were required under the Ontario Laboratory Accreditation program.191

29. In addition, although MDTL was clearly performing a significant number of tests for non-research purposes, the Hospital did not include Dr. Koren as one of its clinical laboratory directors in its licence renewal under the Laboratory and Specimen Collection Centre Licensing Act (Laboratory Licensing Act) until 2009–10.192 With the Hospital’s knowledge, MDTL continued to perform hair tests when it had not been accredited as a clinical laboratory or incorporated in the Hospital’s laboratory licence.

30. Ultimately, MDTL did not participate in its first Ontario Laboratory Accreditation inspection until January 2011 – over a decade after it had begun routinely to offer hair tests to child protection agencies, and more than five years after SickKids became aware of how frequently MDTL was conducting hair tests for child protection agencies.193 The Hospital explained its delay as follows:

When it became apparent that the activities of the MDTL involved a significant non-research (“clinical”) component, steps were taken to align the MDTL with the DPLM to ensure appropriate quality oversight. While well intentioned, the execution of these steps was slow, due in part to assurances from the MDTL of their successful participation in external proficiency testing, as well as an initial focus on improvement of non-analytical policies and procedures between 2005 and 2007. In a world of competing obligations, the activities of the MDTL appeared to be relatively low risk by virtue of their proficiency testing results.

31. This explanation may well address the delay, but it does not justify the result of permitting MDTL to continue offering its hair-testing services in the interim. The Hospital knew that MDTL was carrying out hair tests that could have significant legal consequences. Even if MDTL’s proficiency test results were reliable (which they were not), MDTL’s hair tests were not “relatively low risk.”

4.1 Delay in Improvements to MDTL’s Analytical Methods and Operations

32. Although the decision was made in 2008 not to include MDTL in its midterm self-­assessment under the Ontario Laboratory Accreditation program, the Hospital told the Independent Review that “sights were set on the December, 2010 accreditation inspection instead.” This decision by the Hospital to bring MDTL into the Ontario Laboratory Accreditation framework was critical to the improvements that were ultimately made to MDTL’s analytical methods and operations. It was the push to meet the Ontario Laboratory Accreditation requirements that finally gave MDTL the impetus it needed, first, to stop using ELISA quantitatively and move to gas chromatography–mass spectrometry (GC-MS) confirmation for most adult and child samples, and, second, to develop a quality management system.

33. The transition from ELISA to GC-MS was a slow one. The Independent Review received two accounts of individuals raising with MDTL’s leadership the need for routine GC-MS analysis even before 2005. Dr. Kapur, who was a scientist / consultant toxicologist at SickKids at the time, told the Independent Review that he had a vivid recollection of telling Ms. Klein that MDTL should be routinely confirming its immunoassay results not long after he joined SickKids in 1995 (he could not recall exactly when).

34. Although Ms. Klein could not recall the specific conversation referenced by Dr. Kapur, she said she did speak to Dr. Koren on several occasions between 2003 and 2005 about the importance of moving the Laboratory to routine GC-MS confirmation for its tests for drugs of abuse. According to Ms. Klein, when she emphasized to Dr. Koren the importance of GC-MS or LC-MS confirmation and the urgency in purchasing the equipment, Dr. Koren responded, “We should look into it.” She ultimately attended two courses to pursue the purchase of the necessary equipment.

35. However, when the Independent Review asked Dr. Koren about Ms. Klein’s recollection, he denied that Ms. Klein had raised the issue of GC-MS confirmation with him:

Ms. Klein did not ever raise with me the importance of confirmation of ELISA with GC-MS or LC-MS. To the contrary, Ms. Klein remained a strong proponent of ELISA qualitatively, as well as semi quantitatively, throughout her time at MDTL, for the reasons she has described in her submissions.

36. In any event, whether Ms. Klein raised the issue with Dr. Koren or not, it is clear that MDTL had not purchased a GC-MS instrument to test for drugs of abuse before Ms. Klein’s departure in April 2005. No further steps were taken to pursue the purchase of the equipment until approximately 2008.

37. By 2008, Mr. Gareri had attended four of the annual meetings of the Society of Hair Testing (SoHT) and two clinical toxicology meetings. He learned at those meetings that every other laboratory was using mass spectrometry-based testing (such as GC-MS) and that no one was reporting immunoassay-based results. Having become aware that MDTL was behind other laboratories, in or around 2008 he began pushing for MDTL to purchase a GC-MS instrument that could be used for testing for drugs of abuse.

38. Mr. Gareri also advocated for MDTL to hire a technologist to assist in the development of GC-MS methods for testing for drugs of abuse.194 In a July 18, 2008, email that MDTL produced to the Independent Review, Mr. Gareri wrote as follows to Dr. Koren:

As a side note; beyond what we’ve discussed in terms of increased staff, it is extremely important to develop GCMS methods for not just opiates, but for cocaine, amphetamines and cannabinoids. MS-confirmation is standard in every tox lab I’ve encountered and at some point it will become a very major issue for us if we do not start confirming all our positive results. Regardless of these external factors, it is very important that we eliminate the risk of false positives due to the high stakes involved in our tests. It hasn’t quite worked out having graduate students validate our methods; I still think they should play a role in our future method validation, but we need a dedicated operator working with [Ms. Karaskov] to get routine confirmation analysis happening as soon as possible. I am sure a lack of secondary confirmation of positive results will be a major accreditation issue as well. [Emphasis added.]

39. Dr. Koren responded positively to Mr. Gareri’s attempts to move MDTL toward routine GC-MS confirmation in 2008, agreeing in a responding email to Mr. Gareri that “GC-MS validation is key.”

40. Three months later, in October 2008, MDTL hired Paula Walasek as a research technologist to develop a GC-MS method for testing for drugs of abuse (on a GC-MS instrument that was in the Laboratory for research purposes). However, it was not until March 2009 that MDTL purchased a new GC-MS instrument that was intended to be used exclusively for testing for drugs of abuse.195 The instrument arrived in the Laboratory in May 2009, but it had to be replaced in November 2009 because of its poor performance. From November 2009 onward, MDTL staff continued to work on developing a GC-MS method to test for drugs of abuse on the new instrument.

41. Ultimately, it took approximately two years (from October 2008 to August 2010) for MDTL to develop and validate its GC-MS method to test for drugs of abuse to the level of sensitivity it desired. Although method development on a new GC-MS instrument can often take months, two years was inordinately long.196

42. In any event, by September 2010 MDTL had its GC-MS procedure up and running for drugs of abuse, and by December 2010, it had standard operating procedures and a quality management system in place – just in time for the Ontario Laboratory Accreditation inspection in January 2011. Although deficiencies remained in MDTL’s analytical methods after January 2011, there is no doubt that significant improvements were made. Unfortunately, inadequate and unreliable ELISA tests continued to be used for forensic purposes during the years of the Hospital’s inaction and delay. If the Hospital had acted sooner to bring MDTL into the Ontario Laboratory Accreditation process, many of the improvements that were implemented in the fall of 2010 may well have come into place sooner as well.

5. Lack of Transparency about MDTL’s Methods

43. MDTL reported unconfirmed ELISA results from 2005 to August 2010 – a practice that was inadequate, unreliable, and fell woefully short of internationally recognized forensic standards. MDTL failed to advise the participants in the justice system of its inadequate methods and, as the Independent Review has identified, in several respects the Laboratory also mischaracterized its analytical methods to the scientific community and to users of its test results – thereby making MDTL’s inadequacies all the more difficult to detect.

44. First, researchers at MDTL, including Dr. Koren, published several articles based on data obtained from the MDTL database. As set out in Table 8.1, in those articles the authors indicated that they used ELISA and that positive results were confirmed using GC-MS (or liquid chromatography) when that was not so.

45. The results from the samples described in the articles were not confirmed routinely by GC-MS.197 In fact, between 2005 and 2010, MDTL sent very few samples to a reference laboratory for confirmation. Dr. Koren, who was listed as an author for each of these articles, provided no explanation for why the methods were described in this manner except to say:

Please note that the questions answered by each of these research papers do not relate to the GC/MS versus ELISA comparison, but rather to, for example, maternal/child correlations, timing for the drug to disappear from the hair. GC/MS confirmation was not, in my view, relevant to those conclusions. [Emphasis in original.]

Table 8.1 Methods reported in academic articles

Article

Authors

Methods reported

“Clinical Applications of Hair Testing for Drugs of Abuse – The Canadian Experience” (2000) 107 Forensic Science International 281–8

Julia Klein, Tatyana Karaskov, and Gideon Koren

See page 283:

“After establishing and validating several hair tests during the last decade, we have analysed over 1000 hair samples for different drugs of abuse. Most of the samples were analysed for cocaine and its metabolite – benzoylecgonine.

“… For the screening of the hair either radioimmunoassay (RIA) or enzyme linked immunosorbent assay (ELISA) methodologies were used. The positive results were confirmed using GC-MS.” [Emphasis added.]

“Methamphetamine Detection in Maternal and Neonatal Hair: Implications for Fetal Safety” (2007) 92 Archives of Disease in Children, Fetal & Neonatal Edition 351–5

Facundo Garcia-Bournissen, Ben Rokach, Tatyana Karaskov, Joey Gareri, and Gideon Koren

See page 352:

“The Motherisk Laboratory at the Hospital for Sick Children in Toronto, Canada, routinely receives neonatal and adult hair samples for analysis from various hospitals throughout Canada. These include samples taken from neonates and their corresponding mothers.

“On the basis of clinical suspicion of maternal drug misuse, the testing of hair was requested by either a doctor or the Children’s Aid Societies, with the consent of the subjects or their legal guardians (in the case of children). The tested population included mostly women and their children, but some male adults were also tested, according to the needs of the Children’s Aid Societies that refer the patients.

“Hair samples were analysed following previously self-described methods … Individual drugs were analysed by ELISA, and positive results confirmed by gas chromatography / mass spectrometry.” [Emphasis added.]

epub:

“Cocaine Detection in Maternal and Neonatal Hair: Implications to Fetal Toxicology” (2007) 29(1) Therapeutic Drug Monitoring 71–6

Facundo Garcia-Bournissen, Ben Rokach, Tatyana Karaskov, and Gideon Koren

See page 72:

“The Motherisk Laboratory at the Hospital for Sick Children in Toronto, Canada, routinely receives neonatal and adult hair samples for analysis of cocaine from various hospitals and children’s aid societies throughout Canada.

“Testing of hair is usually requested based on suspicion or evidence of maternal drug abuse by either a physician or the Children’s Aid Societies with consent of the subjects or their legal guardians (in the case of children).

“Hair samples were analysed using methods previously published by the authors … Individual drugs were analyzed by enzyme-linked immunosorbent assay and positive results confirmed by gas chromatography / mass spectrometry with the mass selective detector operating in selective ion monitoring mode as previously published.” [Emphasis added.]

“Contamination of Hair with 3, 4-Methylene Dioxymethamphetamine (Ecstacy) in 2 Young Girls from a ‘Meth Lab’” (2008) 47(2) Clinical Pediatrics 186–8

Facundo Garcia-Bournissen, Farhan Asrar, Zulfikarali Verjee, Tatyana Karaskov, and Gideon Koren

See page 187:

“The presence of MDMA in hair was qualitatively confirmed by high-performance liquid chromatography with diode array ultraviolet detection, with identity confirmation against a drug library (REMEDi HS system; Bio-Rad, Hercules, California).” [Emphasis added.]

“Pharmacokinetics of Disappearance of Cocaine from Hair after Discontinuation of Drug Use” (2009) 189 Forensic Science International 24–7

Facundo Garcia-Bournissen, Monique Moller, M. Nesterneko, Tatyana Karaskov, Gideon Koren

See page 25:

“The Motherisk laboratory at the Hospital for Sick Children in Toronto, Canada routinely receives neonatal and adult hair samples for toxicological analysis from various hospitals and Children’s Aid Societies throughout Canada. The Motherisk Database contains a variety of toxicology results from adults, children and neonatal hair. Consent for analyses is obtained from the subjects or their legal guardian at the time of sampling.

“… Individual drugs were analyzed by ELISA; positive results were routinely confirmed by gas chromatography / mass spectrometry.” [Emphasis added.]

46. First, to state that GC-MS confirmation was “not relevant” to the conclusions drawn in these academic articles is not so. The ELISA results that MDTL generated were preliminary and unconfirmed. To the extent that the research studies relied on the data in MDTL’s database, the methods used to generate the data were “relevant.” In any event, Dr. Koren has indicated that erratum letters have now been sent to the editors of these journals advising them of the errors in the descriptions of the methods used. SickKids has also advised that its research integrity adviser is conducting an inquiry into the concerns raised by these articles. Second, MDTL described itself to the SoHT as carrying out routine GC-MS confirmation when, in fact, it was not.198

47. Third, nowhere in its communications with its customers and with participants in the justice system did MDTL advise that its ELISA tests were preliminary screening tests only. When MDTL made the change to GC-MS in September 2010, it similarly did not make clear the significant change in its methodology.

48. Ultimately, the manner in which MDTL represented its analytical methods to the scientific and legal community and to others created the illusion that MDTL was carrying out hair tests that were adequate and reliable for their intended purposes. In fact, they were not.

6. No Meaningful Oversight of MDTL

49. MDTL was a laboratory within SickKids – one of the most well-known and widely respected children’s hospitals in Canada. The fact that MDTL was housed in such a prominent institution bolstered its reputation and credibility with customers and with the community – the assumption being that such a highly regarded institution would take steps to ensure that the services provided by its laboratories met the standards of excellence that have come to be associated with the SickKids name. Unfortunately, in the case of MDTL, that assumption proved incorrect.

50. The Hospital described oversight of MDTL to the Independent Review as an “evolving concept” as follows:

Given the evolving nature of the MDTL’s research and services, oversight of the lab was a similarly evolving concept, with oversight transitioned over time from its origins in the [Research Institute] to the DPLM as MDTL’s non-research (“clinical”) function became prominent and more widely recognized at the Hospital.

51. Notwithstanding this evolution, the Independent Review has received no evidence to suggest that any department within SickKids took any steps to oversee the Laboratory’s work. Indeed, none of the departments that the Hospital identified as being relevant to MDTL’s oversight exercised any meaningful oversight over MDTL’s operations during the period under review:

  • Motherisk Program: The Independent Review has not obtained any evidence to suggest that the Motherisk Program (of which MDTL was one facet) exercised any separate oversight role with regard to MDTL’s operations. Indeed, Dr. Koren was the director of the Motherisk Program as well as MDTL’s laboratory director. One would not have expected Dr. Koren to play an independent oversight role with respect to his own laboratory.
  • Research Institute: Research laboratories that perform research do so within the Research Institute under the guidance of scientists who are referred to as “principal investigators.” Although principal investigators (including Dr. Koren) reported to the chief of research, the Research Institute’s role was not to exercise any operational or analytical oversight over MDTL as a laboratory – and it did not.199
  • Department of Paediatric Laboratory Medicine: Although there were earlier discussions (in 2005 and 2006) about transitioning MDTL from a research laboratory to a clinical laboratory, MDTL was not formally moved from the Research Institute to DPLM until the fall of 2014.200 In the meantime, however, DPLM also did not exercise meaningful oversight of MDTL’s operations. Its involvement was limited to performing internal audits on MDTL on the basis of select sections of the Ontario Laboratory Accreditation requirements.201 However, these internal audits did not amount to meaningful oversight of MDTL’s operations or analytical or interpretive practices.
  • Division of Clinical Pharmacology and Toxicology (the Division): Although the Division played a role in the oversight of Dr. Koren (who was a clinician within the Division), it was not intended to, and did not, oversee laboratories within the Hospital, including MDTL.202

52. MDTL created an unusual situation for the Hospital. It started as a research laboratory, for which Dr. Koren, as principal investigator, was responsible. It evolved and, by 2005, MDTL was carrying out hair tests on well over a thousand samples each year for child protection cases. However, when the Hospital learned that MDTL was doing tests for non-research purposes, it did not charge any of its programs, departments, or divisions with the responsibility of ensuring that MDTL was subject to appropriate oversight mechanisms. Although the Hospital took steps to bring MDTL into the Ontario Laboratory Accreditation process to ensure that it complied with the relevant clinical licensing and accreditation requirements (albeit far too slowly), it took no steps to ensure that MDTL’s hair tests were fit for their intended purpose or that MDTL had the appropriate equipment, infrastructure, personnel, and expertise to carry out hair tests for drugs of abuse and alcohol markers. Ultimately, none of the programs, departments, or divisions within the Hospital ever took ownership of MDTL as a laboratory, and no clear lines of accountability were ever drawn. The result was inevitable: MDTL effectively slipped through the cracks.

53. The Hospital explained that it relied on MDTL’s clinical accreditation under the Ontario Laboratory Accreditation process as well as its participation in the SoHT proficiency tests. In addition, the Hospital said that DPLM (which was responsible for the Hospital’s clinical laboratories) lacked the expertise to exercise oversight of MDTL’s processes or results:

The participation by MDTL in an international program with the Society of Hair Testing was seen as an indicator of performance based on external standards. Once [Ontario Laboratory Accreditation] accreditation was achieved, the consensus feeling was that proper systems were in place to ensure that the lab was functioning at a high level. Given the fact that hair testing was and is a relatively novel type of test, the DPLM did not have the expertise to assess the MDTL’s processes or results at a more granular level.

54. DPLM’s lack of expertise to conduct meaningful oversight of MDTL should have raised a red flag. The Hospital should have appreciated that, unless it had the expertise to assess the proficiency tests and MDTL’s purported compliance with clinical accreditation standards, it was not doing its job – to provide meaningful oversight. The Hospital should have obtained the requisite expertise to oversee MDTL’s operations or considered stopping the service altogether. It took neither of these steps.

55. My conclusion that SickKids failed to exercise meaningful oversight over MDTL’s work must be considered in the context of the Hospital’s experience with Dr. Charles Smith, a pediatric pathologist who worked at SickKids from July 1980 to September 2005 and who was the first director of the Ontario Pediatric Forensic Pathology Unit (OPFPU), which was housed in the Hospital. On April 25, 2007, the Government of Ontario appointed the Honourable Stephen T. Goudge as commissioner of the Inquiry into Pediatric Forensic Pathology in Ontario. The Hospital was a party with standing at the inquiry. Commissioner Goudge released his report on October 1, 2008 (Goudge Report) – the year in which the Hospital determined that clinical accreditation might not be achieved for MDTL.

56. The Goudge Report is relevant to this Independent Review for at least three reasons. First, Commissioner Goudge determined that Dr. Smith, who was a pediatric pathologist by training, lacked the professional skills necessary for the forensic pathology work that he was performing at the OPFPU at SickKids.203 Commissioner Goudge ultimately recommended that the OPFPU continue as a regional forensic pathology unit located at SickKids, but that “[i]ts director must be a certified forensic pathologist.”204 Even in the face of the Goudge Report, the Hospital failed to appreciate the dangers associated with having a laboratory within the institution that routinely provided a forensic service yet was led by individuals who lacked any forensic training.

57. Second, Commissioner Goudge found that oversight of the OPFPU was wanting, describing a “lack of clarity” in the lines of accountability and oversight as follows:

The 1991 Agreement regarding the OPFPU therefore failed to clearly allocate the responsibility for supervision and oversight. The lines of accountability and oversight were so unclear that the central witnesses each described a different view of the respective roles and obligations of the Chief Coroner, the Chief Forensic Pathologist, and the OPFPU director. This lack of clarity, combined with the fact that no one stepped forward to take responsibility for oversight, resulted in a vacuum where nobody was held to account for the work of the OPFPU.205

58. Although the above paragraph refers to a lack of clarity in the lines of accountability from the OPFPU to the Office of the Chief Coroner for Ontario (which is responsible for death investigations in the province), what Commissioner Goudge describes is not unique to that particular relationship. There was an analogous lack of clarity in the case of MDTL and SickKids. The Hospital failed to draw clear lines of accountability for MDTL’s operations once the Laboratory’s hair tests moved out of the realm of research. None of the other departments within SickKids ever stepped up and took responsibility for oversight of MDTL’s operations.

59. Third, Commissioner Goudge noted the role that SickKids’ reputation played in positioning Dr. Smith as a leading expert in his field:

The position of director of the OPFPU assisted in positioning Dr. Charles Smith to become the leading expert in pediatric forensic pathology, when he lacked the requisite training and qualifications. The mere fact that he came from SickKids, where the OPFPU was located, added significantly to his stature. Yet, in reality, SickKids had no ownership of his forensic pathology work. Thus, it is argued, Ontario got the worst of both worlds – the reputation without the substance on which that reputation should have been based.206

60. The Hospital exercised a similar lack of “ownership” of the work of MDTL. Just as the SickKids name assisted in positioning Dr. Smith to become a leading expert in pediatric forensic pathology, that name likely gave credibility to the work of MDTL as well. In the face of the Goudge Report, the Hospital should have appreciated that others would rely on its reputation and ensured that customers and other users of MDTL’s hair tests – including participants in the justice system – got the “substance” behind that reputation. Again, the Hospital did not do so. It is my expectation that the Hospital will review this Report as part of an important exercise of self-reflection and will develop a detailed plan to ensure that what went wrong at SickKids does not happen again.


Notes

183 Mr. Gareri told the Independent Review that he tried to visit frequent users of MDTL services once a year, sometimes twice, to give this training.

184 The December 2006, December 2007, and December 2009 versions of the “Motherisk Drug Testing Newsletter for Children’s Aid Societies” also indicated that the newsletter was “dedicated to you.”

185 MDTL charged higher rates to parents and other “private” customers.

186 For example, Mr. Gareri attended rounds at MDTL and at the Hospital, as well as the annual conferences of the Society of Hair Testing and other organizations.

187 The previous laboratory manager, Ms. Klein, left the Hospital suddenly in April 2005, so there was no transition between her and Mr. Gareri. Owing to allegations of fraud surrounding the departure of both Ms. Klein and a senior secretary in the Research Institute at the time, the Hospital became concerned about the accessioning of samples and the accounting of money. As a result, Mr. Gareri was charged with implementing a set of pre- and post-analytical procedures.

188 The first interpretation report signed by Mr. Gareri that the Independent Review obtained was dated November 2005.

189 The exception was the Broomfield trial: Although Mr. Gareri testified at the preliminary inquiry in 2007, he said he was not comfortable testifying at the trial, so Dr. Koren testified at the trial in 2009.

190 When the Independent Review asked the Hospital for clarification as to which aspects of the accreditation process might “not be achieved” by MDTL in 2008, it responded that “current Hospital and DPLM leadership do not know the answer to this question.” For his part, Dr. Koren told the Independent Review that “the concern that the accreditation ‘may not be achieved’ reflect[ed] DPLM’s concern and decision to move for 2010 accreditation.”

191 See OLA requirements, version 4.1 (released July 2008), ss II, VI.3.

192 Laboratory and Specimen Collection Centre Licensing Act, RSO 1990, c L.1. Clinical laboratories in Ontario are required to be licensed under the Laboratory Licensing Act. The licence renewals do not list the clinical laboratories within the facility, but do list the clinical laboratory directors. In May 2009, the Hospital wrote to the Ministry of Health and Long-Term Care to add Dr. Koren as a laboratory director for the purposes of the Hospital’s licence, effective June 1, 2009.

193 The next accreditation inspection was scheduled for December 2010 but ultimately took place in January 2011.

194 Until this time, MDTL relied on one its graduate students to work up a GC-MS method for opiates as part of her research for her master of science degree. However, Mr. Gareri recognized that reliance on graduate students to develop and validate GC-MS methods would take too long,

195 Dr. Koren told the Independent Review that the technology to run GC-MS was expensive, and it took time to accumulate adequate funding from MDTL’s revenues.

196 One of the reasons it took so long to implement the GC-MS procedure was because MDTL decided to develop a method that would permit it to test for 17 drugs and metabolites simultaneously.

197 Mr. Gareri told the Independent Review that, although he was not involved in the study, he believed that the samples referred to in the “Contamination of Hair with 3, 4-Methylene Dioxymethamphetamine (Ecstacy) in 2 Young Girls from a ‘Meth Lab’” article were likely tested at SickKids’s “core laboratory” (not MDTL), because the tests referred to in the article were a standard test performed by that laboratory.

198 See Chapter 7.

199 According to the Hospital, such oversight is deferred to the principal investigator. However, the Research Institute did assist in the overhaul of MDTL’s pre- and post-analytical procedures, which were implemented in July 2007.

200 The Hospital told the Independent Review that, in 2013/2014, the executive vice president of clinical programs became aware that Dr. Koren had established a fetal alcohol spectrum disorder classroom within the Toronto District School Board and “became concerned about the operational independence and initiated the transfer of the clinical operations of MDTL to DPLM.” That process was initiated in the late summer/fall of 2014 and was under way when the Court of Appeal’s decision in the Broomfield appeal was released on October 21, 2014.

201 In 2008, DPLM performed a safety audit of MDTL. In 2009, DPLM performed an audit on safety and also on equipment, reagents, and supplies – two sections of the Ontario Laboratory Accreditation requirements. In 2010, DPLM performed an audit on the basis of all 10 sections of the Ontario Laboratory Accreditation requirements (in the lead-up to MDTL’s first accreditation inspection in January 2011). Once MDTL passed its first accreditation inspection, the DPLM’s internal audits in 2012 and 2014 were on only two sections of the Ontario Laboratory Accreditation requirements (organizational structure, personnel, policies, and laboratory management and also on quality management system).

202 Instead, the Division considered laboratory oversight to fall within the purview of the principal investigator (for research laboratories) or DPLM (for clinical laboratories).

203 See, e.g., Ontario, Inquiry into Pediatric Forensic Pathology in Ontario, Report, vol 2 (Toronto: Ministry of the Attorney General, 2008), 116–19 (Commissioner Stephen T. Goudge).

204 Ibid, vol 3, 323 (Recommendation 27).

205 Ibid, vol 2, 92. The 1991 Agreement refers to the agreement that was signed between SickKids and the Ministry of the Solicitor General, establishing the OPFPU.

206 Ibid, vol 3, 321.

Chapter 9
MDTL and Child Protection

1. Introduction

1. Between 2005 and 2015, the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) tested more than 24,000 hair samples for drugs of abuse and alcohol for child protection purposes. The samples represented hair from more than 16,000 different individuals – approximately 54 percent of which tested positive.207

2. In this chapter, I discuss the ways in which MDTL hair test results were used by Ontario’s child protection agencies throughout their work with families. A brief overview of the Child and Family Services Act (CFSA) provides readers with the statutory framework against which I review how the hair test evidence was used in child protection proceedings in Ontario.208

2. Overview of the Child and Family Services Act

2.1 The CFSA’s Purpose

3. In Ontario, child protection services are governed by the Child and Family Services Act.209 The Act’s paramount purpose is “to promote the best interests, protection and well being of children.”210 In addition, the CFSA recognizes that, “while parents may need help in caring for their children,” that help should, wherever possible, support the autonomy and integrity of the family unit, be provided on the basis of mutual consent, and be the least disruptive course of action available and appropriate to help a child.211

4. Promoting the best interests of the child requires a consideration of a variety of factors, including the child’s physical, mental, and emotional needs and level of development; cultural background and religious faith; the importance for the child’s development of a secure place in a positive parental relationship; relationships and emotional ties to family and community; the importance of continuity; the child’s views and wishes; the effect of delay; the degree of risk; and any other relevant circumstances.212 For Aboriginal children, a consideration of their best interests must include “consideration [of] the importance, in recognition of the uniqueness of Indian and native culture, heritage and traditions, of preserving the child’s cultural identity.”213

5. In promoting the autonomy and integrity of the family unit, the CFSA recognizes the expanding nature of family relationships and their evolution to include a range of adult figures in a child’s life. The Act recognizes that the role of “parent” includes a biological parent or any other individual having lawful custody of the child, but does not include a foster parent.214 Relatives and extended family are also broadly defined to include relatives by blood, through a spousal relationship, or by adoption.215 For children who are Aboriginal, extended family includes any member of the child’s band or native community.216

6. Part III of the CFSA specifically governs child protection cases. A child is “in need of protection” if, among other circumstances identified in the Act, the child suffers, or is at risk of suffering, physical, sexual, or emotional harm or neglect, or if the parents are unable to provide care for the child.217

2.2 The CFSA and Child Protection

7. The child-centred approach mandated by the CFSA is implemented by Ontario’s 47 child protection agencies, under the general oversight of the Ministry of Child and Youth Services. The agencies investigate allegations, protect children, provide services to families, care for and supervise children, and place children for adoption.218

8. All persons in Ontario have an obligation to report to a child protection agency when they have reasonable grounds to suspect that a child is or may be “in need of protection.”219 However, children most often come to the attention of an agency through the report of a relative, a school, or a doctor. Once informed about a case where harm or neglect is suspected, the child protection agency uses well-established criteria and designated tools to decide whether, and to what extent, the agency should intervene to protect a child.220 In deciding if and what intervention is appropriate, a child protection worker will interview parents, caregivers, and relatives, as well as community professionals such as teachers, doctors, and social workers. Interventions by a child protection agency may range from supporting the family in the home, to recommending parenting courses or addiction treatment, to the temporary removal of the child from the home, or to a longer or even a permanent removal of the child, including by means of a court disposition of Crown wardship. After assessing the degree of risk to a child, the child protection worker may offer services to the family to keep the child in the home. However, if parents are “temporarily unable to care adequately” for the child, they may by written agreement give the agency temporary care and custody.221 Such agreements usually provide terms and conditions for access by the parent to the child and are for limited periods of time.222 Agencies most commonly resolve child protection concerns by entering into voluntary agreements with parents. In certain circumstances, however, a child protection agency may apprehend a child, with or without a warrant, in order to take the child to a place of safety while the agency initiates court proceedings under Part III of the CFSA.223

9. The first step in the court proceeding will be a hearing to provide for temporary care. Under the legislation this hearing must occur within five days of apprehension,224 which means the parents will often be unrepresented and require an adjournment so they can retain counsel and gather the information necessary to respond to the child protection agency.225 Pending the return of the case, the court will make a temporary order either returning the child to the care of the parents (or other caregivers), with or without supervision by the agency, or order that the agency have temporary care of the child. If the child is placed temporarily with an agency, the child is often placed with a foster family, usually with access by the parents to the child.

10. In child protection proceedings, the court must make two separate determinations, both on the civil standard of a balance of probabilities. First, the court must determine whether the child is in need of protection.226 Second, if the child is in need of protection, the court must then decide which of the following dispositions is in the child’s best interests:227

  • a supervision order placing the child with a parent or other person, subject to supervision by the agency;
  • an order for society wardship bringing the child into the agency’s temporary care for a specified period, which usually results in placement in a foster home;228
  • an order for Crown wardship removing the child from parental care;229 or
  • consecutive orders of society wardship and supervision.

11. An order for supervision or society wardship is subject to a status review hearing.230 A Crown wardship order is also subject to status review, but a parent needs leave of the court to bring such an application if the child has been in the continuous care of the same foster parent or the same custodial person for at least two years.231 No review is available for a child placed for adoption if the Crown ward still lives in a prospective adoptive home.232

12. To advance the CFSA goal of achieving permanency for a child, the child protection framework imposes strict timelines. The Act limits society wardship to 12 months for a child under six years of age, and to 24 months for a child six years or older. These time frames are subject to a six-month extension if it is found to be in the child’s best interests. In keeping with the goal of expeditious disposition, the Family Law Rules anticipate case management and provide a timetable for every step in a child protection case, ending ideally with a disposition within 120 days.233 A court may extend the timetable only if the extension meets the best interests of the child.234 However, the reality is that many child protection cases take much longer as matters are adjourned several times, with temporary orders for care and custody, as the parties work to assemble the necessary information and the family and agency explore resolution.

13. Accordingly, contested child protection proceedings may take two, sometimes three years to reach a final disposition. However, contested cases that proceed to summary judgment motion, or to trial, are by far the minority. The statistics provided by the Ministry of the Attorney General indicate that, between 2005 and 2014, about 10 percent of the approximately 10,000 child protection proceedings each year were scheduled for a summary judgment motion or a trial. That percentage significantly diminished in the last two years to 4.3 percent in 2014.235

3. Expert Evidence

14. Courts must make decisions based on the admissible evidence presented. In child protection cases, the common law rules of evidence, the Evidence Act, the CFSA, and the Family Law Rules govern the admissibility of evidence before the court and how that evidence may be admitted.236

15. As a general rule of evidence, witnesses are prohibited from testifying about their opinions and may testify only to facts about which they have personal knowledge. Expert witnesses, in the absence of any exclusionary rule, are permitted to give opinion evidence as an exception to this general rule, if the witness is properly qualified as an expert and the evidence is both relevant and necessary to assist the court in drawing conclusions from facts.

16. Expert witnesses must have acquired special or peculiar knowledge through training or experience in the subject matter of their opinion. If the expert opinion is based on novel or contested science, or the opinion relies on established science for a novel purpose, the court must be satisfied that the underlying science is reliable for that purpose. In addition to possessing the required expertise, an expert witness must be independent and impartial. The common law has long recognized that an expert witness has a duty to assist the court by providing an objective unbiased opinion.

17. In child protection proceedings, expert evidence will be tendered in an expert report, a copy of which must be provided to each of the parties and filed with the court.237 On September 1, 2011, amendments to the Family Law Rules came into effect, codifying the long-standing common law obligation that experts provide independent unbiased opinion evidence.238 Accordingly, every such expert has a duty

  • to provide opinion evidence that is fair, objective, and non-partisan;
  • to provide opinion evidence that is related only to matters that are within the expert’s area of expertise; and
  • to provide such additional assistance as the court may reasonably require to determine a matter in issue.239

18. These duties prevail over any obligations that an expert may owe to the party that retained or engaged her or him to provide evidence. The Family Law Rules also detail what must be contained in an expert report, including the expert’s qualifications; the instructions provided to the expert; the nature of the opinion sought; the expert’s opinion; the expert’s reasons for the opinion; a list of every document relied on; and a signed acknowledgement of the duty to be fair, objective, and non-partisan.240 The expert’s reasons for the opinion must include a description of the factual assumptions on which the opinion is based and a description of any research that led the expert to form the opinion.241

19. Accordingly, expert opinions such as results and interpretation reports from MDTL must comply with the requirements for expert evidence in order to be admissible in court. However, consistent with the child-centred focus of the legislation, the CFSA mandates less restrictive evidentiary principles than in other proceedings. For example, on adjournments of temporary care hearings, a court may admit, and act on, evidence that the court considers “credible and trustworthy in the circumstances.”242 As a result, arguably, hearsay or even double hearsay “may” be admitted.243

20. In addition, the CFSA permits a court to consider “past conduct” of a person toward any child “if that person is caring for or may care for or have access to a child who is the subject of the proceeding.”244 Further, clause 50(1)(b) of the Act makes admissible any “oral or written statement or report [of past parental conduct] that the court considers relevant to the proceeding,” including those from an earlier proceeding.245

21. It appears that hair test results and interpretations from MDTL were often considered to meet the requirements of expert evidence on summary judgment motions and trials and to be reliable “past conduct” evidence. On adjournments, they were also often accepted as “credible and trustworthy” because the Laboratory’s results were perceived to be objective expert evidence generated from a laboratory housed in the world-renowned Hospital for Sick Children.

4. MDTL Hair Tests and Child Protection Work

22. Child protection agencies work with some of the most vulnerable and marginalized individuals in our society, many of whom lack basic financial and emotional supports. Some turn to drugs and alcohol as a means of self-medication, and some develop an addiction. No matter the emotional bond between parent and child, if a parent’s drug or alcohol abuse creates a risk of neglect or physical, emotional, or sexual harm, the child may be “in need of protection” within the meaning of the Act.

23. In the late 1990s, MDTL began receiving regular requests from child protection agencies asking that hair samples be tested for drugs of abuse. Starting in 2001, MDTL began actively to promote its hair-testing services to child protection agencies through various presentations and seminars.246 The submissions that the Independent Review received indicated that the province’s 47 child protection agencies varied in whether, how often, and for what purposes they used MDTL’s hair test results. However, many child protection agencies described MDTL’s hair tests as a useful tool

  • to determine whether a parent was a chronic alcohol or substance abuser;
  • to identify the substance used and the frequency and quantity of use;
  • to assist in determining whether and to what extent children in the home had been exposed to illicit drugs; and
  • to determine whether substance abuse treatment had been successful and a parent had discontinued use.

24. On the other hand, not all child protection agencies used MDTL’s tests. Some used them only in limited circumstances. More than one of the agencies told the Independent Review that it did not request hair tests if the parent acknowledged drug or alcohol use other than perhaps to test for reduction or cessation of use after treatment.

25. For the most part, child protection agencies reported that hair test results were only one tool used to assess a child’s safety and risk. As one agency noted, “several other factors, such as the state of the home, the demeanor of the parents, the parents’ overall health, the child’s functioning, attachment and development are weighed in decision making.” Another child protection agency put it this way:

[T]he hair-strand test results did not replace the child protection worker’s clinical assessment and judgment of the protection concerns taking into consideration the overall family functioning, and parental capacity.

26. One child protection agency submitted that positive hair test results needed to be supported by evidence “that the parent’s substance abuse is compromising their parenting,” while another agency noted that negative test results meant that children were sometimes returned to their parents.

27. In contrast, submissions from organizations such as the Family Lawyers Association and from individual lawyers who do child protection work described MDTL hair test results as having a significant impact on child protection cases. Their submissions noted that MDTL hair tests were “highly relied on” and a “material factor in the outcome” of cases and were “a significant factor in deciding whether children were returned to their parents, both at case conferences and at trial.” One submission reported that agencies increasingly used hair testing “more and more as the only proof that a parent is not doing drugs.”

28. One lawyer practising child protection described the impact of the MDTL results this way:

A clean test or a negative test can mean the difference between unsupervised or supervised access[,] a return of the child or the agency seeking permanent wardship. Even when all other factors (such as ability to parent, mental health, physical health, community supports) being positive in a parent’s life, courts are reluctant to return a child to a parent in child protection matters where a parent tests positive for drugs and/or results show a high level of use of alcohol and/or marijuana.

29. These submissions also raised the concern that child protection agencies considered MDTL hair test results to be “gospel” or “infallible,” often concluding that the test results established that a parent was lying about drug or alcohol use. A conclusion that the parent had lied about substance use had the potential to affect the tenor of the parent’s relationship with the agency. In addition, such a conclusion could, and sometimes did, reflect adversely on the parent’s credibility in the eyes of the agency, the parent’s own counsel, and the court.

4.1 Use of MDTL Hair Test Results at Various Stages of Proceedings

30. At each stage of a child protection proceeding, MDTL test results were employed in various ways depending on the practice of the child protection agency. As submissions from the Family Lawyers Association and Aboriginal Legal Services of Toronto described it, MDTL tests were used

  • to confirm suspicions of drug and alcohol use;
  • to obtain an accurate level of use;
  • to test a caregiver’s credibility;
  • to monitor levels of drug and alcohol use over time and assess a parent’s compliance with terms and conditions for access to a child;
  • as a term of a court order;
  • as significant evidence of a caregiver’s drug or alcohol use, or the exposure of children to drug use; and
  • to encourage a parent or caregiver to consent to agency intervention, including a temporary care order.

31. In addition, the Family Lawyers Association expressed particular concern about the use of MDTL hair test results in temporary care hearings held after an apprehension. At such hearings, parents are often unrepresented by counsel and have had little time to gather information to respond to allegations of drug and/or alcohol abuse beyond mere denial. In such cases, the Family Lawyers Association said, decisions are often made “on the basis of the agency’s evidence alone.” At that point, a positive hair test may “be a significant and almost incontrovertible piece of evidence” in part because “it affects the tenor and trajectory of the case.” A child may remain in foster care for some time, and at the next hearing the parent will also be faced with “a disadvantageous status quo.”

4.2 The Introduction of MDTL’s Hair Test Results in Court

32. It is noteworthy that child protection agencies regularly introduced MDTL’s hair test evidence in court through affidavit, often by a child protection worker attaching MDTL results and interpretation reports as exhibits. Some workers simply attached a results report (with no interpretation) or noted the fact of a positive test, and the quantitative result with reference to the concentration ranges, in the body of their affidavit. Other affidavits recounted the content of a telephone consultation with an MDTL laboratory counsellor.247 In short, MDTL interpretations were commonly admitted as evidence on the basis of the child protection worker’s information and belief at all stages of child protection proceedings: temporary care hearings, status review hearings, and even sometimes on summary judgment motions.

33. Although these results and interpretations were expert opinion evidence, they were only infrequently recognized or treated as such by child protection counsel, parents’ and children’s counsel, and the court. They were usually introduced into evidence on consent. As a result, the MDTL evidence was rarely tested against the admissibility requirements for expert opinion evidence described above. Accordingly, courts frequently accepted the test results and the concentration ranges as a reliable measure of use with little if any examination of the forensic qualifications of the MDTL representative who communicated those results to the child protection agency, and without any analysis of the adequacy and reliability of the test methodology employed by the Laboratory.

34. In the few cases when MDTL’s expertise or the hair test results were challenged, parents’ counsel rarely produced an opposing expert, and the absence of such an opposing expert was commented on by the courts. In its submission to the Independent Review, the Family Lawyers Association explained that challenging the admissibility of MDTL’s hair test evidence was difficult, in part, for a number of reasons.

  • Lack of supporting documentation. In the experience of their members, MDTL did not routinely provide documentation about its testing methodology or procedure to parties.
  • Lack of information about accreditation. In the experience of their members, MDTL did not disclose that it was not an accredited forensic [laboratory], and did not routinely provide information about the type of accreditation it had achieved and the significance of that accreditation.
  • Lack of funding to test such evidence. A formal challenge to MDTL evidence would require an expert and that funding was not often available through Legal Aid Ontario for expert evidence.248
  • Lack of other facilities or type of testing that is accepted by the courts. Some clients may try to use urine screens to challenge the MDTL hair test results but with mixed success because urine screens can sometimes be seen by the court as less reliable and susceptible to manipulation.

35. As a result, parents were often at a disadvantage. Particularly in summary judgment motions, parents were expected to put their best foot forward, and bald denials of use (or the extent of use) were not generally sufficient to counter the effect of MDTL results. Instead, courts sometimes expected the parent to provide an explanation for the test result or to “rebut” the MDTL evidence.

36. Parents sometimes disputed MDTL’s hair test results, expressing mystification at a positive result. Of particular concern is the class of cases where parents denied any use beyond a particular date, while acknowledging the possibility of passive exposure from their extended family or their living arrangements. If the quantitative result was interpreted as “medium” or “high” based on MDTL’s concentration ranges, MDTL would sometimes advise the child protection worker that passive exposure was extraordinarily unlikely to explain the result. Instead, MDTL sometimes said that passive exposure at those levels could result only from living “in a crack house” or if the parent was “processing drugs.” These conclusions significantly overstate the strength and reliability of the MDTL test results, particularly during the 2005–10 period when MDTL tested for drugs of abuse using enzyme-linked immunosorbent assay (ELISA) tests and did not routinely wash hair samples.

37. Based on all the material, there is support for the views expressed by both the child protection agencies and the lawyers doing child protection work. Among the cases I have reviewed, there are examples where MDTL test results were but a small factor in the outcome and others where it appears that MDTL results were given significant weight. Only a review of the individual facts and evidence in a particular case will reveal whether the outcome was affected by MDTL hair test evidence. Given the more than 9,000 individuals who MDTL tested as positive between 2005 and 2015, a further review is required to consider the extent to which flawed MDTL test results may have affected decisions made about parent–child relationships.


Notes

207 This estimate is based on information provided by the Hospital for Sick Children. Appendix 7 provides the total number of individuals whose hair was tested at the request of an Ontario child protection agency.

208 I use the term “proceedings” here to capture the full scope of a child protection agency’s work with a family, from first contact through to resolution or disposition.

209 Child and Family Services Act (CFSA), RSO 1990, c C.11.

210 CFSA, s 1(1). An overview of the child protection provisions of the Act was prepared for the Inquiry into Pediatric Forensic Pathology in Ontario: see Nicholas Bala and Nico Trocmé, “Child Protection and Pediatric Forensic Pathology,” in Independent Research Studies, Volume 2: Pediatric Forensic Pathology and the Justice System, Kent Roach, director of research (Toronto: Inquiry into Pediatric Forensic Pathology in Ontario, 2008) 43–90. Although prepared in light of concerns about unreliable pathology evidence resulting in wrongful convictions, the study gives historical context to the development of our child protection laws and explains the current child-oriented approach. It is a helpful reference for a more in-depth discussion of the philosophy behind child protection laws.

211 CFSA, s 1(2).

212 See CFSA, s 37(3), for a description of circumstances relevant to a child’s best interests.

213 CFSA, s 37(4). I use the word “Aboriginal” to refer to children who are Indian or native within the meaning of the CFSA. Under subsection 3(1) of the CFSA, “Indian” has the same meaning as in the Indian Act, RSC 1985, c I-5, and “native person” (or child) means a person who is a member of a native community but not a member of a band. Subsection 3(1) of the CFSA further defines native community as a community designated by the minister under section 209 of Part X of the CFSA.

214 CFSA, s 37(1).

215 CFSA, s 3(1).

216 CFSA, s 3(1); in this chapter, the words “family” and “parent” are used in the same way as in the CFSA, and for an Aboriginal child, members of the child’s band or native community.

217 CFSA, s 37(2). See Appendix 15 for the total number of children in care each year from 2005 to 2015, and the number of children adopted.

218 To provide services for families as mandated by the CFSA, child protection agencies must be approved as an “agency” under subsection 8(1), and then designated as a children’s aid society under subsection 15(2) and (3) of the CFSA. For the purpose of this chapter, I will refer to all designated children’s aid societies as “child protection agencies” or “agencies.” I continue to use the term “society wardship” because it is what is used in the Act.

219 CFSA, s 72(1).

220 See, e.g., Ontario Association of Children’s Aid Societies, Ontario Child Welfare Eligibility Spectrum (2006).

221 Part II of the CFSA provides for Voluntary Access to Services, including temporary care agreements. See CFSA, s 29.

222 CFSA, ss 29(10) and 51(6).

223 CFSA, ss 40(2) and 40(7). Once a child has been apprehended, the CFSA requires that, as soon as is practicable, but in any event within five days, the child must be returned to the parents and the matter brought before a court, or the child protection agency must enter into a temporary care agreement with the parents. A place of safety could be a relative’s home, the home of extended family or community, a foster home, a group home, or a safe home as defined in subsection 37(1) of the CFSA.

224 CFSA, s 46(1), Rule 33 of the Family Law Rules, O Reg 114/99.

225 Although not determined at first appearance, in many circumstances legal representation may also be provided for a child (CFSA, s 38). Lawyers are assigned by the province’s Office of the Children’s Lawyer.

226 CFSA, s 37(2).

227 CFSA, s 57. In addition, s 57.1 of the CFSA allows a court to make an order at a protection hearing placing the child in the custody of a person under the Children’s Law Reform Act, RSO 1990, c C.12, without agency supervision; this order may, for example, result in children being placed in the custody of a grandparent.

228 Children who are subject to an order of society wardship will usually see their parents pursuant to an access order made under CFSA, s 58.

229 A Crown wardship order presumptively terminates any order giving the parents access to the child (CFSA, s 59(2)), though a court may still order Crown wardship with access if that is “beneficial and meaningful to the child” and will “not impair the child’s future opportunities for adoption.”

230 CFSA, s 64.

231 CFSA, s 65.1(5).

232 CFSA, s 65.1(9).

233 Rule 33(1), Family Law Rules, O Reg 114/99, made under the Courts of Justice Act, RSO 1990, c C.43.

234 Rule 33(3), Family Law Rules.

235 The number of cases that proceed to summary judgment motion or trial varies significantly depending on the child protection agency and the courthouse. In addition, many of those cases scheduled for a summary judgment motion or trial resolve at the courtroom door. See Appendix 16 for the number of temporary care and custody hearing events, summary judgment motions, and trials scheduled each year from 2005 to 2015.

236 See Evidence Act, RSO 1990, c E.23, and Family Law Rules, O Reg 114/99, made under the Courts of Justice Act, RRO 1990, c C.43. In limited situations, the Charter of Rights and Freedoms may govern the admissibility of evidence in a child protection case.

237 Rules 20.1 and 23(23), Family Law Rules.

238 These amendments were consistent with the 2010 amendments to the Rules of Civil Procedure, RRO 1990, Reg ١٩٤. See rule ٥٣.٠٣ of the Rules of Civil Procedure.

239 Rule 20.1, Family Law Rules.

240 Rule 20.1(6), Family Law Rules.

241 Rule 20.1(6) also requires that an expert also include a summary of and the reasons for any range of opinion.

242 CFSA, s 51(7).

243 Evidentiary principles are also relaxed at bail hearings. See s 518(1)(e) of the Criminal Code, which permits a justice to receive and base a decision on evidence considered credible or trustworthy in the circumstances of each case.

244 See s 50(1)(a) of the CFSA. There are several articles that discuss the meaning of “credible and trustworthy” and “past conduct,” as it relates to evidence in child protection proceedings. For example, see Family and Children’s Services v RO [2006] OJ No 969 (Lexis Nexis); Rollie Thompson, “The Cheshire Cat, or Just His Smile? Evidence Law in Child Protection” (2003) 21 Canadian Family Law Quarterly 319; and Nicholas Bala, “Reforming Ontario’s Child and Family Services Act: Is the Pendulum Swinging Back Too Far?” (1999) 17 Canadian Family Law Quarterly 121. It was not necessary for the Independent Review to explore those issues in light of the focus on whether the hair test results produced by MDTL were reliable for forensic purposes.

245 CFSA, s 50(1). “Oral or written statement or report” includes a transcript, exhibit, or finding, or reasons for a decision in an earlier civil or criminal proceeding.

246 Julia Klein told the Independent Review that she gave approximately 10 presentations to child protection agencies between 2001 and 2004. When Mr. Gareri became laboratory manager in 2005, the frequency increased. Mr. Gareri gave more than 150 presentations to child protection agencies, courts, policing organizations, and other legal organizations between 2005 and 2015, when the Laboratory’s non-research operations were shut down.

247 “Laboratory counsellors” were graduate students and fellows who worked in the Laboratory, often short-term and on a part-time basis, and provided interpretations over the telephone to users of the hair test results.

248 Legal Aid Ontario provided the Independent Review with a copy of its Hair Follicle Memorandum first posted to its website on April 3, 2012. It also maintains a Forensic Science Database created in response to the Goudge Inquiry.

Chapter 10
MDTL and Criminal Proceedings

1. Introduction

1. In addition to their extensive use in child protection proceedings, hair test results from the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) were also used as evidence in criminal proceedings. In fact, the controversy concerning the scientific reliability of hair tests performed by the Laboratory arose in an appeal to the Court of Appeal in a criminal case, R v Broomfield.249

2. In criminal proceedings, hair test results and interpretations generated by the Laboratory have been relied on by the Crown at trial, at guilty pleas, and at sentencing hearings. However, in contrast to hair samples from the over 16,000 individuals tested by MDTL between 2005 and 2015 at the request of child protection agencies, few hair samples were processed by the Laboratory for criminal matters. The information provided to the Independent Review by the Crown Law Office – Criminal identified six cases, including Broomfield, in which MDTL evidence was used by the Crown and a conviction resulted.

3. The differences between criminal and child protection proceedings affected how MDTL results were admitted to court and relied on by the Crown. First, in criminal matters, the Crown has significant pre-hearing disclosure obligations and bears the onus of proving its case beyond a reasonable doubt (in contrast to the disclosure requirements and the balance of probabilities standard in civil proceedings, including child protection matters). Second, generally, the rules of evidence are strictly enforced in criminal matters. Third, and perhaps as a result, MDTL’s results and interpretations were recognized and treated as expert opinion evidence by the criminal justice system.

4. Even so, MDTL’s witnesses were generally accepted as qualified by the parties, and their evidence was not challenged by an opposing expert. The Independent Review is aware of only one criminal case where the defence called an expert to challenge MDTL’s evidence – Broomfield – and on appeal.

2. Broomfield and MDTL Evidence

5. Tamara Broomfield’s child was admitted to a local hospital on August 1, 2005, suffering from seizures. Within hours, the child was transferred to the Hospital for Sick Children (SickKids or the Hospital). On medical examination it was discovered that the child had several broken ribs, in various stages of healing, and a broken wrist. The child was also suffering from a cocaine overdose.

6. The child’s hair was tested by MDTL, first while the child was in hospital at the request of the treatment team, and second, in December 2005, at the request of a child protection agency. Although Ms. Broomfield took the position that her child may have been accidentally exposed to cocaine, MDTL’s interpretations suggested something different. Both Joey Gareri, manager of the MDTL, and Dr. Gideon Koren, the Laboratory’s director, gave evidence that the cocaine levels in the child’s body were so high that they could not have resulted from accidental exposure or ingestion.250 Further, in their opinion, the hair test results revealed that the child had been exposed to cocaine continually over the preceding 14 months.

7. MDTL tested the Broomfield child’s hair in 2005 using its enzyme-linked immunosorbent assay (ELISA) methodology. At trial, Dr. Koren explicitly relied on MDTL’s concentration ranges to explain the high levels of cocaine in the child’s hair:

Q: What’s your database?

A: It’s thousands of cases, at least 1500, thousands. At that time it was about 1500. I thought I brought it. Let me just check. Yes, this is the 40; is very high. To put it into context, 95% of the positive tests we do are below 39, so this is in the five highest measures of positivity that we ever received.

Q: Top 5%?

A: Top 5 percentage, and this is a child of two years. Most of these cases are adult cases that we constructed the database [sic]. I don’t believe we ever [saw] a child with those levels. The same is true, sir, for the level of the Benzoylecgonine, the by-product. They’re very high. [Emphasis added.]

8. Further, although Mr. Gareri’s evidence at the preliminary inquiry implied that MDTL confirmed its ELISA results using gas chromatography–mass spectrometry (GC-MS), Dr. Koren testified that the ELISA method used to test the Broomfield samples was “very sensitive,” “very specific,” and was “a widely accepted method that is published a lot in hundreds of scientific papers and in books and used in courts and in other jurisdictions.”

9. At no point during either the preliminary inquiry or the trial did Dr. Koren or Mr. Gareri inform the court about the limitations of the Laboratory’s hair-testing methodology, specifically that ELISA was a screening test that required confirmation before any of the results could be relied on for forensic purposes.

10. Dr. Koren testified that MDTL was “the reference laboratory in Canada to analyze hair and meconium,” and that the “tests are used all over the world.” He further gave evidence that the Laboratory had “quality assurance protocols” and “tight quality control protocols.” The trial judge accepted Dr. Koren’s evidence, finding that the salient points were “unshaken on cross-examination” and that his evidence was “credible and compelling.”

11. On April 1, 2009, Ms. Broomfield was convicted of aggravated assault and failure to provide the necessities of life in relation to the child’s broken bones. Ms. Broomfield was also convicted of two counts related to the administration of cocaine to the child, in part in reliance on the evidence from MDTL.

3. Opposing Expert Evidence on Broomfield

12. Ms. Broomfield appealed all her convictions but eventually pursued only an appeal of the cocaine-related convictions. On the appeal, she tendered fresh evidence from Dr. Craig Chatterton, the deputy chief toxicologist in the Office of the Chief Medical Examiner of Alberta. Dr. Chatterton’s evidence

  • challenged the methods used to collect and prepare the hair sample tested by MDTL;
  • criticized the methodology used in the analysis of the hair sample; and
  • questioned the validity of the results given in evidence at trial.

13. Specifically, in Dr. Chatterton’s opinion, the results generated by MDTL were not reliable because the Laboratory tested the hair samples using only a screening test (ELISA), and did not confirm the results using GC-MS or liquid chromatography–tandem mass spectrometry (LC-MS/MS). In addition, Dr. Chatterton’s opinion stated that, owing to cross-reactivity, the ELISA screening test could not provide accurate quantitative results and, in any event, the analytical results generated by MDTL should have raised immediate concerns because the concentrations were “so high that they call into question their validity.” Dr. Chatterton also noted that there was no record in the laboratory material that the child’s hair sample had been washed.

14. The Crown responded to Dr. Chatterton’s evidence with fresh opinions from both Dr. Koren and Mr. Gareri.251 In response to Dr. Chatterton’s evidence, both Mr. Gareri and Dr. Koren vigorously defended the Laboratory’s earlier methods (by this time MDTL had transitioned from ELISA to GC-MS to test for most drugs of abuse), relying on the Laboratory’s clinical accreditation and participation in proficiency testing to bolster the reliability of their interpretation in the Broomfield case.

15. In the end, the appeal was argued on the basis of a joint submission by the Crown and the defence. Counsel submitted that the proposed fresh evidence should be reviewed by the court because it was in the interest of justice to do so. Dr. Chatterton’s fresh evidence opinion was to the effect that there was “a genuine controversy among experts about the use of unconfirmed immunoassay testing to support a conclusion of chronic ingestion of cocaine.” The Court of Appeal admitted the fresh evidence from Dr. Chatterton and noted:

No evidence was adduced at trial to challenge the methodology used by the Crown’s expert. The trial judge made her decision unaware of the genuine controversy among the experts about the use of the testing methods relied upon by the Crown expert at trial to found a conclusion of chronic cocaine ingestion, thus, its administration by Ms. Broomfield.252

16. Thus, the Court of Appeal did not conclude that the hair tests carried out by MDTL using ELISA were unreliable, but rather that there was a controversy about those tests and the methodology used. On the basis of the information provided to the Independent Review, there is no doubt that the ELISA tests relied on by MDTL in the Broomfield case were inadequate and unreliable for forensic purposes.

4. Conclusion

17. Public confidence in the justice system demands that any evidence relied on in criminal cases be adequate and reliable for forensic purposes. MDTL’s hair tests did not meet this standard.


Notes

249 R v Broomfield, 2014 ONCA 725.

250 Mr. Gareri testified at the preliminary inquiry in 2007; Dr. Koren testified at the trial in 2009. Both Mr. Gareri and Dr. Koren provided expert reports on the appeal in 2013.

251 The Crown also obtained an opinion from a physician and professor of pharmacology and toxicology at the University of Utah, Dr. Douglas E. Rollins.

252 R v Broomfield, 2014 ONCA 725 at para 12.

Chapter 11
Recommendations

1. Introduction

1. The order in council establishing the Independent Review directed me to conduct a review and provide a report of my findings and recommendations respecting

  1. the adequacy and reliability of the hair-strand drug and alcohol testing methodology utilized by Motherisk between 2005 and 2015 for use as evidence in child protection and criminal proceedings; 
  2. the extent to which the operation of the Motherisk laboratory between 2005 and 2015 was consistent with internationally recognized forensic standards;
  3. other matters related to the operation of the Motherisk laboratory that the Independent Reviewer considers necessary and appropriate to address as a result of her review; and
  4. whether the use of evidence derived from Motherisk’s hair-strand drug and alcohol testing in criminal and child protection proceedings has implications warranting an additional review or process with respect to specific cases or classes of cases and, if so, the nature and extent of any such review or process.

2. For the reasons set out in the previous chapters, I make the following findings:

  1. The hair-strand drug and alcohol testing used by the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) between 2005 and 2015 was inadequate and unreliable for use in child protection and criminal proceedings.
  2. Between 2005 and 2015, MDTL operated in a manner that did not meet internationally recognized forensic standards.
  3. The Hospital for Sick Children did not provide meaningful oversight over MDTL.
  4. The use of MDTL hair-testing evidence in child protection and criminal proceedings has serious implications for the fairness of those proceedings and warrants an additional review.

3. My recommendations follow from these findings.253

2. Recommendations for a Second Review – A Commissioner and a
Review and Resource Centre

4. In addition to the several criminal cases potentially affected by MDTL evidence, child protection agencies obtained positive MDTL results for more than 9,000 individuals.254 Given the large number of cases potentially affected, the Province of Ontario, as expeditiously as possible, should establish a Second Review of those individual cases or classes of cases that may have been affected by MDTL’s flawed hair-testing methodology. Only an additional review will be able to address the extent to which flawed MDTL test results may have affected decisions about parent–child relationships or criminal responsibility.

5. This Second Review should be carried out by a Commissioner appointed to undertake that task and to lead what I will refer to in these recommendations as a Review and Resource Centre (RRC). The RRC will provide efficient, comprehensive access to information, legal advice, counselling, support, assessments, alternative dispute resolution services, and other ancillary resources, including any appropriate scientific testing. With the benefit of these resources, and with the advice of the RRC, individuals will be best able to decide whether to pursue remedies and, if they do, what they should be.

2.1 The Commissioner

6. I recommend that the Province of Ontario appoint a Commissioner under the Public Inquiries Act to conduct the Second Review. The Commissioner will be responsible for the work plan and priorities of the Review and Resource Centre.

7. The Commissioner must be, and must be seen to be, independent of the Province and all participants in the child protection system. I recommend that the Commissioner be a sitting or retired judge with appropriate knowledge and experience to assume this initiative. Such an appointment would convey the importance of the challenges to be addressed and, I anticipate, would encourage the co-operation of the various participants in the system.

8. I am not recommending that the Province of Ontario call a public inquiry. However, I am recommending that the Commissioner be given the tools to conduct a meaningful review. The Commissioner’s authority would include the power to summons witnesses as well as documents and other items relevant to the subject matter of the Second Review. In addition, the Commissioner would have the ability to access material appropriate to the criminal or child protection context – including court files, child protection files, exhibits, and transcripts – in order to fulfill the role of assessing and triaging cases. The Child and Family Services Act (CFSA) quite properly places a premium on maintaining the confidentiality of records relating to child protection proceedings, court files, exhibits, court transcripts, child protection files, and adoption records.255 I recommend that the Province of Ontario take whatever steps are necessary to ensure that the Commissioner can access all these relevant records in an efficient manner while ensuring the confidentiality of those records appropriate to the circumstances.

9. Based on my consultations, including roundtables, I anticipate that the Commissioner will need approximately three years to discharge the functions of the Second Review. That period will provide an opportunity to identify affected cases, guide the participants through the process, and provide appropriate supports and resources.

2.2 The Review and Resource Centre

10. Parents, children, young adults, siblings, and adoptive parents involved in past child protection proceedings who may have been affected by MDTL test results must be provided with the resources necessary to permit them to make an informed decision about any steps that may be available and appropriate. Accordingly, the Commissioner will need the resources of the Review and Resource Centre as the centrepiece of the Province’s compassionate response to those vulnerable persons who may have been further marginalized by the flawed MDTL test results.256

11. The RRC will provide resources that include information on whether an MDTL test was obtained in a particular proceeding; a child protection file review; counselling services; legal advice; appropriate assessments; and alternative dispute resolution services. The RRC will assess an individual’s need for such services and either provide support directly or facilitate ready access to it.

12. The Commissioner’s mandate will touch individuals who live in all parts of the province. Although the RRC may necessarily be physically based in a large municipal centre, it may well need to travel to or have a physical presence in other parts of the province as well.257 It will also need working relationships with service providers across the province to ensure that, to the greatest extent possible, its resources are made available efficiently to Ontario residents in their communities. To assist in this endeavour, the RRC should make effective use of an online presence and telephone communication service, including an immediately available toll-free number and e-mail address answered by well-trained and knowledgeable staff.

2.2.1 Counselling Services

13. It will be extremely difficult for affected individuals to learn how MDTL’s inadequacies may have altered their family relationships. Regardless of whether a hair test result played a material role in the outcome of their proceeding, I anticipate that the information in this Report will cause emotional challenges, particularly for individuals in a vulnerable position. In addition, it will take the Commissioner some time to assess all the cases, some of which will be extremely complex. Children and youth may well face different and quite specific challenges, including not knowing whether a test from MDTL was obtained or if it affected their case. This Report, and the Second Review, will cause difficult and painful issues to resurface.

14. For that reason, I recommend that the RRC provide appropriate counselling assistance to anyone who believes that she or he may have been affected by an MDTL test result. It will be necessary to determine the type of individual or group counselling that will meet each person’s needs and to put individuals in touch with qualified professionals in their communities or online. I envision that the cost of counselling services will form part of the Second Review budget and that those services will be available expeditiously to those in need. The process must preserve the privacy of those who choose to access these services and maintain client–counsellor confidentiality.

2.2.2 Legal Resources

15. I recommend that the RRC offer appropriate legal advice to affected individuals. For complicated cases, or those bound for court, it may be more appropriate for individuals to retain independent lawyers rather than having RRC-appointed counsel as their advocates. The RRC should be able to fund independent lawyers for individuals, including children, where it considers the circumstances appropriate. I recommend that the RRC receive sufficient resources to meet this need.

16. To receive funding, I recommend that, in all but exceptional circumstances, the individual be required to select a lawyer from a roster maintained by the RRC. To appear on the roster, lawyers should receive training from the RRC, demonstrate experience and expertise in the relevant disciplines,258 and agree to work on a tariff model to be established by the RRC. These steps are necessary to ensure the efficiency and effectiveness of the legal services.

17. Where the RRC does not extend funding, individuals should still have the right to apply to Legal Aid Ontario for a certificate to fund a legal challenge.

2.2.3 Alternative Dispute Resolution Services

18. Given the challenges presented by many of the cases, it will be necessary for the RRC to consider and to help craft innovative solutions. In many instances, alternative dispute resolution (ADR), including mediation, may provide appropriate outcomes and achieve expeditious resolutions concerning family relationships that are in the current best interests of the children. The ADR process need not mirror the ADR provisions of the Child and Family Services Act;259 rather, it should be designed to meet the particular needs of the individual cases. When designing the model, the Commissioner may wish to consider various options and to retain qualified mediators from the appropriate professional organizations who are experienced in recognizing and addressing any power imbalances among the participants.

19. As a component of the ADR system, the RRC should be able to organize and pay for fresh, expeditious, and focused parenting assessments to determine if a parent currently has the capacity to have a relationship with a child as well as the potential parameters of such a relationship. In addition, in-house counsellors or social workers could assist in compiling case histories to expedite the assessment of appropriate remedies.

2.3 Scope of the Second Review and Notice about Potentially Affected Persons

20. A Second Review that examines every child protection proceeding in which an MDTL test was obtained would not be efficient, effective, or necessary. My conclusion that MDTL’s hair-strand drug- and alcohol-testing methodology between 2005 and 2015 was neither adequate nor reliable for use as evidence does not mean that every child protection case with an MDTL-tested hair sample reached the wrong result. In the vast majority of cases, other evidence would have amply supported the decision or disposition.

21. In some cases, however, an unreliable MDTL test result may have made a material contribution to the outcome of the case. Unravelling how important the test result or interpretation report was in any particular proceeding would be a complicated forensic exercise. Embarking on this exercise for every one of the thousands of cases with an MDTL hair test result between 2005 and 2015 would be a formidable, time-consuming, expensive, and impractical exercise that would not achieve the desired expeditious and just result.

22. Even if a particular MDTL hair test was inaccurate and played a material role in a proceeding, the court would still be required to determine the current best interests of the child before granting any remedies. The Child and Family Services Act mandates that its primary purpose is to promote and protect the best interests of children. The Second Review must make the best interests of children its primary consideration.260

23. For these reasons, I am not recommending that the Second Review examine every child protection proceeding where an MDTL test result was obtained. Instead, I recommend that a potentially affected person would initiate an individual review by contacting the RRC.261 The Commissioner should, nevertheless, have the discretion on a case-by-case basis to commence a review for a class of cases or for an individual case, including at the request of some other person or agency, if the Commissioner considers it appropriate to do so.

24. Linked closely to the issues of scope and the initiation of an individual review is the question of providing notice about the Second Review to potentially affected persons. In an ideal world, the Second Review would send direct, confidential notice of its creation and mandate to all potentially affected persons. However, based on the information provided by both local child protection agencies and members of the family law bar, the reality is that this wide-scale distribution would prove impossible in practice, given the amount of outdated contact information available. As well, individual notification would be prohibitively expensive and largely ineffective.

25. However, the Commissioner will develop appropriate criteria to determine whether and in what circumstances individual notice should be provided to those affected by the MDTL test results. In particular, I recommend that any parent who had an MDTL test and has a child who is no longer in her or his care, and where that child has not yet been placed for adoption, or the adoption has broken down, should receive personal notice of the Second Review.

26. Similarly, the Commissioner will determine whether, what, and how notice should be given to affected children, including those who are Crown wards or who have been adopted. I recommend that the Commissioner work with children and youth, including youth who have experienced the child protection system, to ensure that their voices, both individually and collectively, are heard throughout the Second Review.

27. The Commissioner should provide widespread public notice of the Second Review in the manner that is most likely to come to the attention of those persons potentially affected by MDTL tests. There are a variety of ways this notice could be delivered, and the RRC should make appropriate use of its website, the media, social media, youth chat and help lines, and other social-networking mechanisms and organizations that serve the community. It will be for the Commissioner to determine the most effective and efficient methods for notifying those affected.

2.4 High-Priority Cases

28. It will take time to establish the Second Review. However, some high-priority cases cannot wait for a thoughtful and deliberate construction of this new agency.

29. First, with the support of the Ministry of Children and Youth Services, child protection agencies should immediately identify any cases involving MDTL hair-testing results that remain open and where a child has not yet been placed for adoption. In such circumstances, child protection agencies should contact the parents or their lawyers to advise them of the potentially flawed hair test results and the creation of the Second Review. Child protection agencies should also assess these cases without regard to MDTL test results unless and until those results are confirmed, if they can be. In addition, child protection agencies should provide a complete copy of the unredacted file to the RRC as soon as possible. This process must be expedited and be given the highest priority following the release of this Report

30. Second, the Commissioner should consider those current cases in which there was an MDTL test result and, in addition, a term of an operative order requiring the parent to comply with ongoing hair testing. It may be appropriate for such cases to return to court for removal or variation of that term.

31. Third, child protection agencies and counsel should ensure that MDTL test results play no role in any case going forward. Whether or not an individual proceeding is reviewed by the Second Review, no person or organization should rely on any MDTL test result for any purpose in any current or future proceeding.

2.5 Children Who Have Been Adopted

32. If an affected child has been placed for adoption (particularly without an openness order) in part because of flawed evidence from MDTL, the currently available range of remedies is legislatively limited. I recommend that the Commissioner consult broadly with affected persons, including youth, adoptive parents, and other caregivers, as well as with experts in the area, in an effort to develop approaches to resolution based on the best interests of the child. The Commissioner will have this opportunity for broad consultation and to ascertain the parameters of the problem, including any questions of public policy, once the circumstances of individual cases or classes of cases have been identified by the Second Review.

33. I recommend that the Commissioner also consider whether, in appropriate circumstances, birth parents or other affected individuals should be provided the ability to file information on the adoption registry or in the child protection file. That information would refer to the MDTL controversy and the possible role of flawed hair tests in affecting the outcome of the child protection proceeding and would become available to affected children who access the registry.

2.6 Resources

34. The Province of Ontario’s investment in the Second Review will help families and individuals who may have been harmed by MDTL’s flawed hair tests by providing the resources to streamline individual cases to an expeditious and just resolution. Such a process will also reduce the potential number of individuals who would otherwise seek immediate redress through the formal adversarial court system and further burden already stretched institutions. Only a sufficiently resourced infrastructure will have the ability to provide meaningful assistance and fulfill its mandate.

35. In addition, child protection agencies, the Office of the Children’s Lawyer, Legal Aid Ontario, and the court system will all be concerned with the resource implications of a Second Review. Although they will all be steadfast in their commitment to ensure that the justice system fairly and expeditiously responds to any cases seeking to restore a parent–child relationship on the basis that a previous decision was tainted by flawed MDTL evidence, they will also be concerned that their resources will be strained and unable to manage current caseloads if large numbers of cases are revisited. Where these institutions require additional resources and oversight to deal with the implications of this Report and the Second Review, I recommend that the Province of Ontario make such resources available.

3. Additional Recommendations Regarding Criminal Cases

36. As described in Chapter 10, MDTL test results were obtained in far fewer criminal proceedings than in child protection proceedings.262 That said, the few individuals who may have been affected by an MDTL test result in a criminal proceeding may well find the content of this Report distressing. I recommend that any such person be permitted to access the services of the RRC, including for the purpose of determining what, if any, steps they may wish to take.

37. In cases in which an individual seeks to set aside convictions based on alleged errors in the Laboratory’s work, I recommend that the Crown Law Office – Criminal should assist in expediting the convicted person’s access to the Court of Appeal and in facilitating a determination of the substantive issues in the cases, unencumbered by unnecessary procedural impediments. Such assistance should include

  1. consenting to defence applications for an extension of time within which to appeal;
  2. working toward agreement with the defence on evidentiary or procedural protocols for applications to extend the time within which to appeal or for introducing fresh evidence on appeal or respecting the appeal itself; or
  3. narrowing the issues that need to be resolved by the Court.

38. The Province of Ontario, either directly or through Legal Aid Ontario, should adequately fund potentially affected persons to access legal and forensic opinions and advice for the purpose of pursuing either an appeal or a new trial or any other legal remedy, including a section 696.1 Criminal Code application to set aside an existing conviction. In appropriate cases, this funding should extend to obtaining an expert opinion to determine whether or not there is merit to such an appeal or application.

4. Financial Compensation

39. The order in council establishing my review did not provide me with a mandate to make recommendations about individual compensation. It permitted me to consider the manner in which hair-strand drug- and alcohol-testing evidence has been used in various legal proceedings but instructed me not to report on any individual cases that are, have been, or may be the subject of child protection or criminal investigations or proceedings. Moreover, the Independent Review is not an inquiry and has not had the benefit of legal argument on the issues of financial compensation.

40. For these reasons, I do not make any recommendations concerning compensation. Questions of any monetary compensation are left to the Second Review, the Province of Ontario, or civil litigation.

5. Recommendation for the Hospital for Sick Children

41. Seven years ago, the Honourable Justice Stephen T. Goudge concluded an Inquiry into Pediatric Forensic Pathology in Ontario.263 That inquiry focused on the flawed pathology evidence of Dr. Charles Smith, once a highly regarded pathologist at the Hospital. Dr. Smith’s testimony led to several wrongful convictions.

42. In his report, Commissioner Goudge assessed “the practice and oversight of pediatric forensic pathology in Ontario” and concluded there were serious and systemic shortcomings in both accountability and oversight. The Goudge Report provided comprehensive recommendations in an attempt to ensure that no one would have to endure the type of harm that can result from flawed science and its acceptance into evidence in the judicial system. The recommendations Commissioner Goudge put forth were far-reaching, with the intention of professionalizing Ontario’s pathology system and providing for oversight and governance of forensic pathology in this province.

43. The citizens of Ontario are justifiably proud of the Hospital for Sick Children, which is one of the world’s leading children’s hospitals. I am confident that the Hospital will reflect deeply on what went wrong at MDTL and within its own institution, particularly because it appears that the Hospital has not adapted and implemented the lessons to be learned from the Goudge Report. There should never need to be another review or inquiry into how the Hospital, its professional staff, or its programs interact with the justice system.

44. In particular, the Hospital for Sick Children should carefully examine its programs to determine which, if any, are offering forensic services, and it should exercise meaningful oversight over any such services. The Hospital must ensure that appropriate standards and training are in place for any of its staff who testify in court as experts. Ultimately, the Hospital will be held as accountable for the quality of any forensic services it provides as it is for the quality of its clinical services.

6. Recommendations for the Justice System

45. In Chapters 9 and 10, I consider the manner in which hair-strand drug and alcohol testing has been used in child protection and criminal proceedings.264 The use of forensic science in court proceedings poses many challenges. As a first step, the scientific evidence that is proffered must itself be accurate and reliable. I have concluded that the MDTL hair test evidence was neither.

46. Commissioner Goudge’s recommendations arising from his review of forensic pathology apply with equal force to the science of forensic toxicology as it relates to hair tests. Of particular relevance are his many recommendations aimed at reducing the risks of an expert opinion being misunderstood: for example, his emphasis on the importance of forensic training, the careful use of terminology and plain language in opinion evidence, transparency, and clear communication of those areas where there are limitations on or controversy concerning the opinion.

47. The Goudge Report provides recommendations for the courts. Those recommendations include the judicial responsibility to set the parameters of the area of expertise and to ensure threshold reliability of both the witness and the science. Following the inquiry’s recommendations, the National Judicial Institute, which is a judge-led organization, increased its science and expert evidence education for judges and made available on its website a Science Manual for Canadian Judges.265 I too recommend ongoing and expanded judicial education regarding expert evidence, particularly expert scientific evidence.

48. I also recommend increased education for members of the bar. Counsel must be increasingly vigilant to ensure that justice is not undermined by the use of flawed forensic evidence. Counsel should ensure that they understand the scope and limitations of a forensic toxicologist’s (or any expert’s) expertise and opinions. They should exercise care to seek to qualify an expert properly and to set precise parameters on that expert’s area of expertise. Counsel must not frame questions that invite forensic experts to stray outside their expertise or the outer boundaries of the science. Lawyers should be vigilant through their questions about the use of terminology that may breed misunderstanding or misinterpretation.

49. I commend Commissioner Goudge’s extensive recommendations to all participants in the justice system. Those recommendations continue to provide crucial guidance on how the legal community must use science in the interests of justice.


Notes

253 The order in council focused my mandate on the Motherisk Drug Testing Laboratory (MDTL or the Laboratory) and its testing methodologies. Accordingly, it is not appropriate for me to address or make recommendations about any other hair-testing laboratories in Ontario. I have no information concerning the manner in which other laboratories perform or interpret hair tests. In addition, in April 2015 the Hospital for Sick Children (SickKids or the Hospital) permanently closed MDTL for all non-research purposes. For this reason, there is no need to make recommendations about improvements that might have brought the Laboratory’s hair-strand testing methodology into compliance with internationally recognized forensic standards.

254 This number is based on data provided to the Independent Review by the Hospital.

255 See Child and Family Services Act (CFSA), RSO 1990, c C.11, Part VII; s 45(10), ss 161–5.

256 The resources outlined in this section, modified to the particular case and context, would also be available where appropriate in criminal cases.

257 A table reflecting the information provided by MDTL, showing the locations of the child protection agencies and the extent that each agency used its hair-testing services, is included as Appendix 17.

258 It may be that lawyers who are members of a panel established by the Office of the Children’s Lawyer would be well suited to this role.

259 See, for example, ss 20.2(1), 145.2(7), and 153.1(10).

260 This recommendation is consistent with the United Nations Convention on the Rights of the Child, where ­Article 3 states that, “in all actions concerning children, whether undertaken by public or private social welfare institutions, courts of law, administrative authorities or legislative bodies, the best interests of the child shall be a primary consideration.” Article 12 of the Convention also speaks to children’s right to be heard in all matters affecting them and states that their views must be given due weight. In particular, it states that children shall be provided with the opportunity to be heard in any judicial and administrative proceeding affecting them, “either directly, or through a representative or an appropriate body.” In my opinion, together with the CFSA, the Convention should inform the process and content of the Second Review.

261 An affected person would also be at liberty to bring her or his own proceeding in court.

262 The information provided to the Independent Review by the Crown Law Office – Criminal identified six cases, including Broomfield, in which MDTL evidence was used by the Crown and resulted in a conviction.

263 Ontario, Inquiry into Pediatric Forensic Pathology in Ontario, Report, 4v (Toronto: Ministry of the Attorney General, 2008) (Commissioner Stephen T. Goudge).

264 My mandate does not extend to other proceedings such as custody and access proceedings.

265 National Judicial Institute, Science Manual for Canadian Judges (Ottawa, 2013), online: https://www.nji-inm.ca/index.cfm/publications.

Appendices

Appendix 1
Order in Council, November 26, 2014

Appendix 1: Order in Council, November 26, 2014
Appendix 1: Order in Council, November 26, 2014
Appendix 1: Order in Council, November 26, 2014
Appendix 1: Order in Council, November 26, 2014
Appendix 1: Order in Council, November 26, 2014

Appendix 2
Order in Council, April 22, 2015

Appendix 2: Order in Council, April 22, 2015
Appendix 2: Order in Council, April 22, 2015
Appendix 2: Order in Council, April 22, 2015
Appendix 2: Order in Council, April 22, 2015
Appendix 2: Order in Council, April 22, 2015
Appendix 2: Order in Council, April 22, 2015

Appendix 3
Meetings with Organizations and Roundtables

Catholic Children’s Aid Society of Toronto

Centre of Forensic Sciences

Children’s Aid Society of Toronto

Family Lawyers Association

Hospital for Sick Children

Institute for Quality Management in Healthcare

Institute of Environmental Science & Research Ltd. – Forensic Toxicology

Legal Aid Ontario

Ministry of the Attorney General

Ministry of the Attorney General, Crown Law Office – Criminal

Ministry of Children and Youth Services

Ministry of Health and Long-Term Care

Motherisk Drug Testing Laboratory staff

Office of the Children’s Lawyer

Ontario Association of Children’s Aid Societies

Ontario Forensic Pathology Service

Provincial Advocate for Children and Youth

Victorian Institute of Forensic Medicine


Child Protection Roundtable Participants

Nicholas Bala, Professor, Faculty of Law, Queen’s University
Lorne Glass, Glass & Associates
The Honourable Justice Heather L. Katarynych, Ontario Court of Justice
Shelley Kierstead, Assistant Professor, Osgoode Law School
Kristina Reitmeier, Lawyer, Children’s Aid Society of Toronto
Bruce Rivers, Executive Director, Covenant House, Toronto
The Honourable Justice Anthony W.J. Sullivan, Ontario Court of Justice

Appendix 4
Submissions Received

Aboriginal Legal Services of Toronto

The Advocates’ Society

The Association in Defence of the Wrongly Convicted (AIDWYC)

Judith Birchall, Lawyer

Bruce Grey Child & Family Services

Catholic Children’s Aid Society of Hamilton

Catholic Children’s Aid Society of Toronto

Chatham-Kent Children’s Services

Children’s Aid Society of Algoma

Children’s Aid Society of Guelph-Wellington

Children’s Aid Society of Haldimand and Norfolk

Children’s Aid Society of Hamilton

Children’s Aid Society of London and Middlesex

Children’s Aid Society of the District of Nipissing and Parry Sound

Children’s Aid Society of the Districts of Sudbury and Manitoulin

Children’s Aid Society of Toronto

Kimberly Costa, Lawyer

The County & District Law Presidents’ Association

Criminal Lawyers’ Association

Defence for Children International – Canada

Dynacare

Marlys Edwardh, Lawyer

Elizabeth Fry Society

Family and Children’s Services Niagara

Family and Children’s Services of Frontenac, Lennox and Addington

Family & Children’s Services St. Thomas & Elgin

Family & Children’s Services of the Waterloo Region

Family Lawyers Association

Highland Shores Children’s Aid

Hospital for Sick Children

Huron-Perth Children’s Aid Society

The Innocence Project, Osgoode Law School

Jewish Family & Child Service of Greater Toronto

Kawartha Haliburton Children’s Aid Society

Kenora-Rainy River Districts Child & Family Services

Lynn Kirwin, Lawyer

Members of the Public – 16 individual submissions

Native Child and Family Services of Toronto

Melanie O’Neill, Lawyer

Office of the Children’s Lawyer

Office of the Provincial Advocate for Children & Youth

Dr. Nancy Olivieri

Ontario Association of Chiefs of Police

Ontario Association of Children’s Aid Societies

Martin J. Prost, Lawyer

W. Gerald Punnett, Lawyer

Peter Rutland, Lawyer

Sarnia-Lambton Children’s Aid Society

Simcoe Muskoka Child, Youth and Family Services

Cara Valiquette, Lawyer

Viaguard Accu-Metrics

York Region Children’s Aid Society

York Region Law Association Child Protection Best Practices Committee

Appendix 5
Reported Child Protection Cases Reviewed

Children’s Aid Society of Durham (County) v R(R), 2005 CarswellOnt 385, [2005]
OJ No 360

Children’s Aid Society of Halton (Region) v N (RR), 2008 ONCJ 95

Children’s Aid Society of Northumberland v P(D), 2008 CarswellOnt 2984, [2009]
OJ No 2047

Children’s Aid Society of Toronto v B(Y), 2008 ONCJ 800

Children & Family Services of York Region v C(H), 2009 CarswellOnt 4994, [2009]
OJ No 3527

Children’s Aid Society of Hamilton v V(L), 2009 CarswellOnt 1904, [2009] OJ No 1468

Children’s Aid Society of Simcoe (County) v L(S), 2009 CarswellOnt 772, 174 ACWS
(3d) 1087

Children’s Aid Society of Toronto v A (RE), 2009 CarswellOnt 6027, [2009] OJ No 4167

Children’s Aid Society of Toronto v A (RE), 2009 CarswellOnt 8380

Children’s Aid Society of Toronto v H (T), 2009 ONCJ 282

Children’s Aid Society of Toronto v L (T), 2009 ONCJ 788

Children’s Aid Society of Toronto v S (D), 2009 CarswellOnt 6725, [2009] OJ No 4605

Children’s Aid Society of Waterloo (Regional Municipality) v A (LJA), 2009 ONCJ 226

Family & Children’s Services of Guelph & Wellington County v B (K), 2009 ONCJ 765

Native Child & Family Services of Toronto v C (D), 2009 CarswellOnt 8998

Catholic Children’s Aid Society of Toronto v V (A), 2010 ONCJ 657

Children & Family Services for York Region v B (T), 2010 ONSC 7047

Children’s Aid Society of Algoma v M (K), 2010 ONCJ 762

Children’s Aid Society of Algoma v W (D), 2010 ONCJ 761

Children’s Aid Society of Halton (Region) v M (KC), 2010 ONCJ 743

Children’s Aid Society of Toronto v L (T), 2010 ONSC 1376

Children’s Aid Society of Waterloo (Regional Municipality) v T (PDA), 2010 ONCJ 739

Family & Children’s Services of Lennox & Addington v W (S), 2010 ONSC 2585

Native Child & Family Services of Toronto v C (D), 2010 ONSC 1038

Native Child & Family Services of Toronto v D (T), 2010 ONCJ 409

Catholic Children’s Aid Society of Hamilton v P (CR), 2011 ONSC 2056

Children’s Aid Society of Algoma v A (L), 2011 ONCJ 92

Children’s Aid Society of Algoma v G (L), 2011 ONCJ 392

Children’s Aid Society of Algoma v L (R), 2011 ONCJ 678

Children’s Aid Society of Niagara Region v B (T), 2011 ONSC 2702

Children’s Aid Society of Toronto v E (RH), 2011 ONCJ 650

Children’s Aid Society of Toronto v G (T), 2011 ONCJ 625

Children’s Aid Society of Toronto v H (T), 2011 CarswellOnt 16014, [2013] WDFL 1359

Children’s Aid Society of Waterloo (Regional Municipality) v M (JL), 2011 ONCJ 734

Catholic Children’s Aid Society of Toronto v M (M), 2012 ONCJ 369

Children’s Aid Society of Simcoe (County) v W (T), 2012 ONSC 3635

Windsor-Essex Children’s Aid Society v B (T), 2012, ONCJ 588

Children’s Aid Society of Algoma v S (R), 2013 ONCJ 688

Children’s Aid Society of Brant v B (J), 2013 ONSC 4059

Children’s Aid Society of London and Middlesex v W (A), 2013 ONSC 4638

Children’s Aid Society of Toronto v B (D), 2013 ONCJ 405

Catholic Children’s Aid Society of Toronto v D (A), 2014 ONCJ 490

Chatham-Kent Children’s Services v K (A), 2014 ONCJ 224

Children’s Aid Society of the Niagara Region v P (L), 2014 ONSC 4914

Children’s Aid Society of Toronto v B, 2014 ONCJ 486

Children’s Aid Society of Toronto v B, 2014 ONCJ 90

Children’s Aid Society of Toronto v J (C), 2014 ONCJ 221

Windsor-Essex Children’s Aid Society v B (T), 2014 ONCJ 239

Appendix 6
Further Reading on the Science of Hair Testing

International Forensic Toxicology Guidelines

The Society of Forensic Toxicologists, Inc & The American Academy of Forensic Sciences, Toxicology Section. “Forensic Toxicology Laboratory Guidelines” (2002).

The Society of Forensic Toxicologists, Inc & the American Academy of Forensic Sciences, Toxicology Section. “Forensic Toxicology Laboratory Guidelines – 2006 Version” (2006), online: www.soft-tox.org/files/Guidelines_2006_Final.pdf.

The United Kingdom and Ireland Association of Forensic Toxicologists, “Forensic Toxicology Laboratory Guidelines (2010)” (2010) 50 Science and Justice 166–76. Online www.ukiaft.co.uk/system/files/pdf/UKIAFT%20Guidelines%202010.pdf.

Hair-Testing Guidelines

Society of Hair Testing. “Statement of the Society of Hair Testing Concerning the Examination of Drugs in Human Hair” (1997) 84 Forensic Science International 3–6.

United Nations Drug Control Programme. Guidelines for Testing Drugs under International Control in Hair, Sweat and Saliva (2001).

Society of Hair Testing. “Recommendations for Hair Testing in Forensic Cases” (2004) 145 Forensic Science International 83–4.

Kintz, P., ed. “Consensus of the Society of Hair Testing on Hair Testing for Chronic Excessive Alcohol Consumption, 2009” (2010) 176 Forensic Science International 2.

Cooper, G.A.A., R. Kronstrand, P. Kintz. “Society of Hair Testing Guidelines for Drug Testing in Hair” (2012) 218 Forensic Science International 20–4.

Kintz, P., ed. “Consensus of the Society of Hair Testing on Hair Testing for Chronic Excessive Alcohol Consumption 2011” (2012) 218 Forensic Science International 2.

United Nations Drug Control Programme. Guidelines for Testing Drugs Under International Control in Hair, Sweat and Saliva (2014). Online www.unodc.org/documents/scientific/ST_NAR_30_Rev.3_Hair_Sweat_and_Oral_Fluid.pdf.

Kintz, P., ed. “2014 Consensus for the Use of Alcohol Markers in Hair for Assessment of Both Abstinence and Chronic Excessive Alcohol Consumption” (2015) 249 Forensic Science International A1-A2.

General Information

Harkey, M.R. “Anatomy and Physiology of Hair” (1993) 63(1–3) Forensic Science International 9–18.

Huestis, M.A. “Judicial Acceptance of Hair Tests for Substances of Abuse in the United States Courts: Scientific, Forensic, and Ethical Aspects” (1996) 18 Therapeutic Drug Monitoring 456–9.

Spiehler, V. “Hair Analysis by Immunological Methods from the Beginning to 2000” (2000) 107 Forensic Science International 249–59.

Romano, G., N. Barbera, and I. Lombardo. “Hair Testing for Drugs of Abuse: Evaluation of External Cocaine Contamination and Risk of False Positives” (2001) 123(2–3) Forensic Science International 119–29.

Penders, J., and A. Verstraete. “Laboratory Guidelines and Standards in Clinical and Forensic Toxicology” (2006) 11 Accreditation and Quality Assurance 284–90.

Kintz, P., ed. Analytical and Practical Aspects of Drugs Testing in Hair (Boca Raton: CRC Press, 2007).

Cooper, G.A.A., M. Moeller, R. Kronstrand. “Current Status of Accreditation for Drug Testing in Hair” (2008) 176 Forensic Science International 9–12.

Cooper, G.A.A. “Hair Testing Is Taking Root” (2011) 48 Annals of Clinical Biochemistry 516–30.

LeBeau, M.A., M.A. Montgomery, and J.D. Brewer. “The Role of Variations in Growth Rate and Sample Collection on Interpreting Results of Segmental Analyses of Hair” (2011) 210(1-3) Forensic Science International 110–16.

Buffoli, B., F. Rinaldi, M. Labanca, E. Sorbellini, A. Trink, E. Guanziroli, R. Rezzani, L.F. Rodella. “The Human Hair: From Anatomy to Physiology” (2014) 53 International Journal of Dermatology 331–41.

Kintz, P., A. Salomone, and M. Vincenti, eds. Hair Analysis in Clinical and Forensic Toxicology (Amsterdam: Elsevier, 2015).

Drug Exposure in Children

Gray, T., and M. Huestis. “Bioanalytical procedures for monitoring in utero drug exposure” (2007) 388 Analytical and Bioanalytical Chemistry 1455–65.

Farst, K., J.A.R. Meyer, T.M. Bird, L. James, J.M. Robbins. “Hair Drug Testing of Children Suspected of Exposure to the Manufacture of Methamphetamine” (2011) 18 Journal of Forensic and Legal Medicine 110–14.

Bassindale, T. “Quantitative Analysis of Methamphetamine in Hair of Children Removed from Clandestine Laboratories – Evidence of Passive Exposure?” (2012) 219 Forensic Science International 179–82.

Chatterton, C., and P. Kintz. “Hair Analysis to Demonstrate Administration of Amitriptyline, Temazepam, Tramadol and Dihydrocodeine to a Child in a Case of Kidnap and False Imprisonment” (2014) 23 Journal of Forensic and Legal Medicine 26–31.

Pichini, S., O. Garcia-Algar, A.T. Avarez, M. Mercadal, C. Mortali, M. Gottardi, F. Svaizer, R. Pacifici. “Pediatric Exposure to Drugs of Abuse by Hair Testing: Monitoring 15 Years of Evolution in Spain” (2014) 11 International Journal of Environmental Research and Public Health 8267–75.

Kintz, P. “Contribution of in utero Drug Exposure When Interpreting Hair Result in Young Children” (2015) 249 Forensic Science International 314–17.

Wang, X., and O.H. Drummer. “Review. Interpretation of Drug Presence in the Hair of Children” (2015) 257 Forensic Science International 458–72.

Dose-Response

Henderson, G.L., M.R. Harkey,C Zhou, R.T. Jones, P Jacob. “Incorporation of Isotopically Labeled Cocaine into Human Hair. 1. Dose-Response Relationships.” (1996) 20(1) Journal of Analytical Toxicology 1–12.

Ditton, J., G.A.A. Cooper, K.S. Scott, D.L. Allen, J.S. Oliver, and I.D. Smith. “Hair Testing for ‘Ecstasy’ (MDMA) in Volunteer Scottish Drug Users” (2000) 5 Addiction Biology 207–13.

Paterson, S., R. Cordero, M. McPhillips, and S. Carman. “Interindividual Dose / Concentration Relationship for Methadone in Hair” (2003) 27 Journal of Analytical Toxicology 20–3.

Scheidweiler, K.B., E.J. Cone, E.T. Moolchan, and M.A. Huestis. “Dose-related Distribution of Codeine, Cocaine, and Metabolites into Human Hair Following Controlled Oral Codeine and Subcutaneous Cocaine Administration” (2005) 313 Journal of Pharmacology and Experimental Therapeutics 909–15.

Stability of Drugs in Hair

Felli, M., S. Martello, R. Marsili, M. Chiarotti. “Disappearance of Cocaine from Human Hair after Abstinence” (2005) 154 Forensic Science International 96–8.

Shen, M., P. Xiang, Y. Sun, and B. Shen. “Disappearance of 6-Acetylmorphine, Morphine and Codeine from Human Scalp Hair after Discontinuation of Opiate Abuse” (2013) 227 Forensic Science International 64–8.

Favretto, D., M. Tucci, A., Monaldi, S.D. Ferrara, G. Miolo. “A Study on Photodegradation of Methadone, EDDP, and Other Drugs of Abuse in Hair Exposed to Controlled UVB Radiation” (2014) 6 Drug Testing Analysis 78–84.

Suwannachom, N., T. Thananchai, A. Junkuy, T.E. O’Brien, and P. Sribanditmongkol. “Duration of Detection of Methamphetamine in Hair after Abstinence” (2015) 254 Forensic Science International 80–6.

Alcohol Markers

Yegles, M., A. Labarthe, V. Auwärter, S. Hartwig, H. Vater, R. Wennig, F. Pragst. “Comparison of Ethyl Glucuronide and Fatty Acid Ethyl Ester Concentrations in Hair of Alcoholics, Social Drinkers and Teetotallers” (2004) 145 Forensic Science International 167–73.

De Giovanni, N., G. Donadio, M. Chiarotti. “Ethanol Contamination Leads to Fatty Acid Ethyl Esters in Hair Samples” (2008) 32 Journal of Analytical Toxicology 156–9.

Pirro, V., D. Di Corci, S. Pellegrino, M. Vincenti, B. Sciutteri, A. Salomone. “A Study of Distribution of Ethyl Glucuronide in Different Keratin Matrices” (2011) Forensic Science International 271–7.

Pragst, F. “Interpretation Problems in a Forensic Case of Abstinence Determination Using Alcohol Markers in Hair” (2012) 217 Forensic Science International e4–e7.

Schräder, J., M. Rothe, F. Pragst. “Ethyl Glucuronide Concentrations in Beard Hair after a Single Alcohol Dose: Evidence for Incorporation in Hair Root” (2012) 126 International Journal of Legal Medicine 791–9.

Pianta, A., B. Liniger, M.R. Baumgartner. “Ethyl Glucuronide in Scalp and Non-head Hair: An Intra-individual Comparison” (2013) 48 Alcohol and Alcoholism 295–302.

Kintz, P. “Testing for Ethanol Markers in Hair: Discrepancies after Simultaneous Quantification of Ethyl Glucuronide and Fatty Acid Ethyl Esters” (2014) 243 Forensic Science International 44–6.

Drug Incorporation Mechanisms

Henderson, G.L. “Mechanisms of Drug Incorporation into Hair” (1993) 63 Forensic Science International 19–29.

Rollins, D.E., D.G. Wilkins, and G. Krueger. “Models for Studying the Cellular Processes and Barriers to the Incorporation of Drugs into Hair” (1995) 154 NIDA Research Monograph 235–44.

Joseph, E.R., T.P. Su, E.J. Cone. “In vitro Binding Studies of Drugs to Hair: Influence of Melanin and Lipids on Cocaine Binding to Caucasoid and Africoid Hair” (1996) 20 Journal of Analytical Toxicology 338–44.

Pötsch, L., and M.R. Moeller. “On Pathways for Small Molecules into and out of Human Hair Fibres” (1996) 41 Journal of Forensic Science 121–5.

Kronstrand, R., and K. Scott. “Drug Incorporation into Hair,” in P. Kintz, ed, Analytical and Practical Aspects of Drug Testing in Hair (Boca Raton: CRC Press, 2007) 1–23.

Lee, S., E. Han, E. Kim, H. Choi, H. Chung, S.M. Oh, Y.M., Yun, S.H. Jwa, K.H. Chung. “Simultaneous Quantification of Opiates and Effect of Pigmentation on Its Deposition in Hair” (2010) 33 Archives of Pharmacological Research 1805–11.

Polettini, A., E.J. Cone, D.A. Gorelick, M.A. Huestis. “Incorporation of Methamphetamine and Amphetamine in Human Hair Following Controlled Oral Methamphetamine Administration” (2012) 72 Analytica Chimica Acta 35–43.

Influence of Cosmetic Treatments and Washing

Balíková, M.A., and V. Habrdová. “Hair Analysis for Opiates: Evaluation of Washing and Incubation Procedures” (2003) 789 Journal of Chromatography B, Analytical Technologies in the Biomedical and Life Sciences 93–100.

Binz, T.M., M.R. Baumgartner, T. Kraemer. “The Influence of Cleansing Shampoos on Ethyl Glucuronide Concentration in Hair Analyzed with an Optimized and Validated LC-MS/MS Method” (2014) 244 Forensic Science International 20–4.

Kidwell, D.A., F.P. Smith, and A.R. Shepherd. “Ethnic Hair Care Products May Increase False Positives in Hair Drug Testing” (2015) 257 Forensic Science International 160–4.

Cocaine Interpretation

Schaffer, M., V. Hill, and T. Cairns. “Hair Analysis for Cocaine: The Requirement for Effective Wash Procedures and Effects of Drug Concentration and Hair Porosity in Contamination and Decontamination” (2005) 29(5) Journal of Analytical Toxicology 319–26.

Pragst, F., and M.A. Balíková. “State of the Art in Hair Analysis for the Detection of Drug and Alcohol Abuse” (2006) 370 Clinical Chimica Acta 17–49.

Stout, P.R., J.D. Ropero-Miller, M.R. Baylor, J.M. Mitchell. “External Contamination of Hair with Cocaine: Evaluation of External Cocaine Contamination and Development of Performance-Testing Materials” (2006) 30 Journal of Analytical Toxicology 490–500.

LeBeau, M.A., and M.A. Montgomery. “Considerations on the Utility of Hair Analysis for Cocaine” (2009) 33(6) Journal of Analytical Toxicology 343–4.

Pragst, F., H. Sachs, P. Kintz. “Hair Analysis for Cocaine Continues to be a Valuable Tool in Forensic and Clinical Toxicology” (2010) 34 Journal of Analytical Toxicology 354–6.

Ropero-Miller, J.D., M.A. Huestis, and P.R. Stout. “Cocaine Analytes in Human Hair: Evaluation of Concentration Ratios in Different Cocaine Sources, Drug-User Populations and Surface-Contaminated Specimens” (2012) 36(6) Journal of Analytical Toxicology 390–8.

Morris-Kukoski, C.L., M.A. Montgomery, and R.I. Hammer. “Analysis of Extensively Washed Hair from Cocaine Users and Drug Chemists to Establish New Reporting Criteria” (2014) 38 Journal of Analytical Toxicology 628–36.

Appendix 7
Hair-Testing Data, 2005–15, Provided by the Hospital for Sick Children

Entity

Samples tested

Samples with negative test results

Samples with positive test results

Individuals tested

Individuals with negative test results

Individuals with positive test results

Child protection agencies 2005-2007

4,199

1,387

2,812

3,262

1,049

2,213

Child protection agencies 2007-2015

13,611

5,556

8,055

9,443

4,698

4,745

Third-party collectors 2007-2015

6,390

2,903

3,487

5,758

2,712

3,046

Totals

24,200

9,846

14,354

18,463

8,459

10,004

Percentages

40.7%

59.3%

45.8%

54.2%

Caution to Readers:
These totals are based on data provided by the Hospital for Sick Children. The Hospital advises that the data it provided is an estimate only. It has not been verified by the Independent Review. The Hospital told the Independent Review that third-party collectors arranged for hair testing on behalf of child protection agencies, and in some circumstances on behalf of private individuals. At least 1200 individuals are likely double-counted because their hair was collected by the third-party collectors OSAD and DriverCheck on behalf of Ontario child protection agencies. In addition, some individuals whose hair was collected by third-party collectors resided in Nova Scotia and New Brunswick. Accordingly, the actual number of individuals tested at the request of Ontario child protection agencies is likely lower than as set out in the totals line. Please note that the Hospital collected and reported data for 2005–07 and 2007–15 separately.

Hair-Testing Data, 2005–7, Provided by the Hospital for Sick Children

Child protection agency

Samples tested

Samples with negative test results

Samples with positive test results

Individuals tested

Individuals with negative test results

Individuals with positive test results

Algoma CAS

13

1

12

13

1

12

Brant CAS

10

7

3

10

7

3

Bruce County CAS

169

68

101

141

54

87

CAS Hamilton

1391

431

960

993

285

708

CAS Toronto

637

200

437

513

157

356

CCAS Hamilton

658

215

443

514

168

346

CCAS Toronto

611

200

411

454

143

311

Chatham-Kent

4

2

2

4

2

2

Dufferin CAS

2

1

1

2

1

1

Durham CAS

187

42

145

153

36

117

Frontenac, Lennox and Addington

10

2

8

9

2

7

Guelph and Wellington

3

1

2

3

1

2

Haldimand-Norfolk CAS

14

8

6

13

8

5

Halton

3

2

1

3

2

1

Highland Shores

118

34

84

91

19

72

Huron-Perth

1

0

1

1

0

1

Jewish Family and Child Service

10

5

5

7

3

4

Kawartha-Haliburton

8

2

6

8

2

6

Kenora-Rainy River Districts

8

2

6

8

2

6

Lanark, Leeds and Grenville

31

3

28

31

3

28

London and Middlesex

5

1

4

5

1

4

Native Toronto

16

1

15

12

1

11

Family and CS of Niagara

8

4

4

8

4

4

North Eastern Ontario

11

5

6

9

3

6

Ottawa CAS

11

2

9

10

2

8

Peel CAS

37

17

20

33

15

18

FCS Renfrew County

1

0

1

1

0

1

Simcoe & Muskoka CAS

39

11

28

34

10

24

Sudbury and Manitoulin CAS

1

0

1

1

0

1

Thunder Bay CAS

4

3

1

4

3

1

FACS Waterloo Region

11

7

4

11

7

4

Windsor Essex CAS

62

35

27

61

34

27

York Region CAS

105

75

30

102

73

29

Totals for 2005-2007

4,199

1,387

2,812

3,262

1,049

2,213

Percentages

33.0%

67.0%

32.2%

67.8%

Caution to Readers:
This data was provided by the Hospital for Sick Children. The Hospital advises that the data it provided is an estimate only. It has not been verified by the Independent Review.

Hair-Testing Data, 2007–15, Provided by the Hospital for Sick Children

Child protection agency

Samples tested

Samples with negative test results

Samples with positive test results

Individuals tested

Individuals with negative test results

Individuals with positive test results

Anishinaabe Abinoojii FS

24

4

20

23

4

19

Brant Family & CS

12

7

5

12

8

4

Bruce Grey C&FS

1367

678

689

775

487

288

CAS County of Renfrew

3

1

2

1

1

0

CAS Guelph & Wellington

26

13

13

26

14

12

CAS Haldimand-Norfolk

26

9

17

22

8

14

CAS Nipissing & Parry Sound

75

12

63

68

15

53

CAS of Algoma

1395

536

859

842

430

412

CAS of Hamilton

1522

487

1035

1057

419

638

CAS of London and Middlesex

4

1

3

4

2

2

CAS of Ottawa

104

50

54

88

45

43

CAS of Oxford County

27

13

14

25

15

10

CAS Region of Peel

275

124

151

226

111

115

CAS Sarnia & Lambton

6

1

5

6

2

4

CAS St Thomas and Elgin

39

17

22

31

18

13

CAS Stormont, Dundas, Glengary

1

1

0

0

0

0

CAS Sudbury & Manitoulin

494

153

341

351

127

224

CAS Thunder Bay

51

17

34

46

19

27

CAS Toronto

1580

667

913

1131

571

560

Catholic CAS Hamilton

1056

406

650

603

301

302

Catholic CAS Toronto

1351

610

741

917

514

403

Chatham-Kent CS

113

51

62

72

39

33

Dilico Anishinabek Family Care

246

67

179

205

67

138

Dufferin Child & Family Services

6

1

5

1

1

0

Durham CAS

559

242

317

380

209

171

F&CS Waterloo Region

67

44

23

65

45

20

Family & Children’s Services Niagara

87

53

34

84

54

30

Frontenac, Lennox & Addington

43

16

27

39

19

20

Halton CAS

156

76

80

128

70

58

Highland Shores CAS

332

144

188

250

123

127

Huron-Perth CAS

18

13

5

18

14

4

Jewish Family & Child Service of GTA

47

14

33

34

15

19

Kawartha-Haliburton CAS

25

12

13

21

12

9

Kenora Rainy River C&FS

42

20

22

25

16

9

Kunuwanimano C&FS

1

1

0

1

1

0

Lanark, Leeds & Grenville

180

76

104

136

69

67

Native C&F Service Toronto

256

93

163

182

87

95

NorthEastern Ont F & CS

817

295

522

604

266

338

Payukotayno

1

1

0

1

1

0

Simcoe Muskoka CY & FS

311

140

171

256

128

128

Weechi-it-te-win FS

112

28

84

83

27

56

Windsor-Essex CAS

112

58

54

111

60

51

York Region CAS

642

304

338

493

264

229

Totals for 2007-2015

13,611

5,556

8,055

9,443

4,698

4,745

Percentages

40.8%

59.2%

49.8%

50.2%

Third-party collector

Samples tested

Samples with negative test results

Samples with positive test results

Individuals tested

Individuals with negative test results

Individuals with positive test results

Bayshore Healthcare

2230

981

1249

1980

892

1088

Bayshore Home Health

2679

1217

1462

2564

1197

1367

DriverCheck Inc.

697

339

358

557

294

263

OSAD Inc.

784

366

418

657

329

328

Totals for 2007-2015

6,390

2,903

3,487

5,758

2,712

3,046

Percentages

45.4%

54.6%

47.1%

52.9%

Caution to Readers:
This data was provided by the Hospital for Sick Children. The Hospital advises that the data it provided is an estimate only. It has not been verified by the Independent Review. The Hospital told the Independent Review that third-party collectors arranged for hair testing on behalf of child protection agencies, and in some circumstances on behalf of private individuals. At least 1200 individuals are likely double-counted because their hair was collected by the third-party collectors OSAD and DriverCheck on behalf of Ontario child protection agencies. In addition, some individuals whose hair was collected by third-party collectors resided in Nova Scotia and New Brunswick. Accordingly, the actual number of individuals tested at the request of Ontario child protection agencies is likely lower than as set out in the totals line.

Appendix 8
Drug Concentration Ranges

8a Ranges of Concentration in Hair – Earliest Version

8a Ranges of Concentration in Hair – Earliest Version

8b Ranges of Drug Concentration in Adult Hair – March 2006

8b Ranges of Drug Concentration in Adult Hair – March 2006

8c Ranges of Drug Concentration in Infant Hair – March 2006

8c Ranges of Drug Concentration in Infant Hair – March 2006

8d Ranges of Drug Concentration in Adult Hair – August 2011

8d Ranges of Drug Concentration in Adult Hair – August 2011

8e Interpretation of Drug Detection in Child Hair – August 2011

8e Interpretation of Drug Detection in Child Hair – August 2011

Appendix 9
MDTL, The Use of Hair Testing to Establish Illicit Drug Use

Appendix 9: MDTL, The Use of Hair Testing to Establish Illicit Drug Use
Appendix 9: MDTL, The Use of Hair Testing to Establish Illicit Drug Use
Appendix 9: MDTL, The Use of Hair Testing to Establish Illicit Drug Use
Appendix 9: MDTL, The Use of Hair Testing to Establish Illicit Drug Use

Appendix 10

Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers
Appendix 10: A Review of Substance Abuse Monitoring in a Social Services Context: A Primer for Child Protection Workers

Appendix 11
MDTL Report Templates

11a Results Report Template – redacted and annotated by MDTL

11a Results Report Template – redacted and annotated by MDTL

11b Interpretation Report Templates – redacted – May 2007

11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007
11b Interpretation Report Templates – redacted – May 2007

11c Interpretation Report Template – Adult – July 2011

11c Interpretation Report Template – Adult – July 2011
11c Interpretation Report Template – Adult – July 2011
11c Interpretation Report Template – Adult – July 2011
11c Interpretation Report Template – Adult – July 2011
11c Interpretation Report Template – Adult – July 2011
11c Interpretation Report Template – Adult – July 2011

11d Interpretation Report Template – Child & Neonate – June 2014

11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014
11d Interpretation Report Template – Child & Neonate – June 2014

Appendix 12
MDTL Interpretation Guidelines

12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009

12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009
12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009
12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009
12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009
12a MDTL Interpretation Guidelines (Child Hair Samples) – November 2009

12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009

12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009
12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009
12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009
12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009
12b MDTL Interpretation Guidelines (Adult Hair Samples) – November 2009

Appendix 13
Alcohol Marker Concentration Ranges

13a Ranges of FAEE Concentration in Adult Hair – Earliest Version

13a Ranges of FAEE Concentration in Adult Hair – Earliest Version

13b Ranges of FAEE Concentration in Adult Scalp Hair – July 2007

13b Ranges of FAEE Concentration in Adult Scalp Hair – July 2007

13c Ranges of FAEE Concentration in Adult Scalp Hair – March 2010

13c Ranges of FAEE Concentration in Adult Scalp Hair – March 2010
13c Ranges of FAEE Concentration in Adult Scalp Hair – March 2010

13d Interpretation of Alcohol Hair Analysis, Adult Scalp Hair – March 2011

13d Interpretation of Alcohol Hair Analysis, Adult Scalp Hair – March 2011

13e Interpretation of Alcohol Hair Analysis, Adult Scalp Hair – May 2011

13e Interpretation of Alcohol Hair Analysis, Adult Scalp Hair – May 2011

Appendix 14
MDTL Letter to Clients, October 15, 2010

Appendix 14: MDTL Letter to Clients, October 15, 2010
Appendix 14: MDTL Letter to Clients, October 15, 2010

Appendix 15
Number of Adoptions from Individuals in Care, 2005–14

2005-6

2006-7

2007-8

2008-9

2009-10

2010-11

2011-12

2012-13

2013-14

Total in care

18,500

18,800

18,000

17,900

17,416

17,000

17,060

16,358

15,569

Temporary care

3800

3800

3400

3100

3000

3247

3361

3255

N/A

Society wards

1700

1700

1500

1400

1240

1109

1143

1063

970

Crown wards

9400

9400

9300

9400

9125

8516

7915

7486

6980

Adoptions completed

823

803

748

819

993

971

837

830

923

Available adoption homes

1439

1400

1500

1461

1600

1543

1582

1707

1811

Note: These numbers were provided by the Ontario Association of Children’s Aid Societies.

Appendix 16
Child Protection Proceedings Statistics, 2005-15

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

Jan–Mar
2015

Proceedings received*

13,163

12,685

11,403

11,046

10,528

10,535

10,388

10,336

9,725

9,268

2,241

Temporary care and custody hearing events scheduled

1,716**

2,172

2,009

2,125

1,802

2,022

2,192

2,109

2,038

1,959

597

Summary judgment motions scheduled

388

477

449

472

512

573

625

626

533

258

5

% of Proceedings with summary judgment motions

3.0

3.8

3.9

4.3

4.9

5.4

6.0

6.1

5.5

2.8

0.2

Trials scheduled

600

579

538

542

525

492

462

413

331

137

4

% of Proceedings with trials

4.6

4.6

4.7

4.9

5.0

4.7

4.4

4.0

3.4

1.5

1.2

* This row contains the total number of proceedings received by the Ontario Court of Justice and the Ontario Superior Court of Justice.

** This number represents those events scheduled between April and December, 2005.

Note: The data in this table was collected by the Ministry of the Attorney General.

Appendix 17
Number of Individuals Tested by MDTL, 2005–15, Organized by
Ontario Child Protection Agency

Name of organization

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Total

Anishinaabe Abinoojii FS

1

8

10

3

1

23

Brant Family & CS

3

5

3

6

1

1

1

2

22

Bruce Grey C&FS

17

72

91

99

115

114

112

155

67

63

3

908

CAS Algoma

35

152

100

104

104

105

124

126

5

855

CAS County of Renfrew

1

1

2

CAS Guelph & Wellington

1

2

5

2

8

3

2

5

1

29

CAS Haldimand-Norfolk

6

3

4

4

4

5

3

1

3

2

35

CAS Hamilton

458

329

466

383

215

165

25

3

3

1

2048

CAS Nipissing & Parry Sound

1

1

6

20

10

29

1

68

CAS of London & Middlesex

5

3

1

9

CAS of Ottawa

1

9

3

5

9

22

19

13

11

6

98

CAS of Oxford County

2

2

1

2

9

4

3

2

25

CAS Region of Peel

15

14

11

15

22

61

29

47

28

17

259

CAS Sarnia & Lambton

1

1

2

2

6

CAS St. Thomas & Elgin

2

11

8

3

1

6

31

CAS Stormont, Dundas, Glengary

CAS Sudbury & Manitoulin

1

1

6

65

63

92

111

11

350

CAS Thunder Bay

14

2

2

8

4

1

14

5

50

CAS Toronto

188

216

248

194

190

163

95

106

106

123

6

1635

CCAS Hamilton

205

223

219

153

106

40

46

51

39

32

2

1116

CCAS Toronto

161

201

227

212

165

116

114

78

38

50

5

1367

Chatham-Kent CS

1

3

4

10

13

29

15

1

76

Dilico Anishinabek Family Care

6

34

25

52

24

24

38

2

205

Dufferin Child & Family Service

2

1

3

Durham CAS

62

61

78

71

88

20

22

30

45

50

1

528

F&CS Waterloo Region

4

3

13

11

20

15

5

2

2

75

Family & Children’s Services Niagara

2

2

4

4

8

21

12

14

8

13

4

92

Frontenac, Lennox & Addington

3

5

4

2

8

10

7

5

3

47

Halton CAS

2

1

6

19

13

16

30

18

6

17

1

129

Highland Shores CAS

20

48

55

44

59

29

32

24

10

18

339

Huron-Perth CAS

1

2

3

4

3

2

3

1

19

Jewish Family & Child Service

5

2

6

7

4

6

1

1

6

2

1

41

Kawartha-Haliburton CAS

5

2

1

3

3

2

2

4

3

4

29

Kenora-Rainy River C&FS

4

6

4

2

2

4

3

4

4

33

Kunuwanimano C&FS

1

1

Lanark, Leeds & Grenville

5

14

30

53

21

11

13

7

5

6

165

Native C&F Service Toronto

2

6

10

23

20

24

15

24

36

30

4

194

NorthEastern Ont F & CS

3

26

34

23

34

53

75

157

203

4

612

Payukotayno

1

1

Simcoe & Muskoka CY&FS

14

9

36

42

61

23

29

27

22

25

288

Weechi-it-te-win FS

16

11

8

13

12

22

1

83

Windsor-Essex CAS

16

32

33

30

21

6

3

2

7

17

167

York Region CAS

38

48

30

35

50

75

88

77

82

71

1

595

TOTAL

1240

1321

1666

1626

1394

1162

1042

1029

1012

1113

53

12658

Third-party collector

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Total

Bayshore Healthcare

163

270

276

227

260

285

278

212

9

1980

Bayshore Home Health

5

292

482

432

541

485

174

144

5

2560

DriverCheck Inc.

60

197

149

149

2

557

OSAD Inc.

23

74

121

260

179

657

TOTAL

191

636

879

919

1040

967

601

505

16

5754

Caution to Readers:
This data was provided by the Hospital for Sick Children. The Hospital advises that the data it provided is an estimate only. It has not been verified by the Independent Review. The Hospital told the Independent Review that third-party collectors arranged for hair testing on behalf of child protection agencies, and in some circumstances on behalf of private individuals. At least 1200 individuals are likely double-counted because their hair was collected by the third-party collectors OSAD and DriverCheck on behalf of Ontario child protection agencies. In addition, some individuals whose hair was collected by third-party collectors resided in Nova Scotia and New Brunswick. Accordingly, the actual number of individuals tested at the request of Ontario child protection agencies is likely lower than as set out in the totals line.

Independent Reviewer and Review Staff



INDEPENDENT REVIEWER
The Honourable Susan E. Lang

Lead Counsel to the Independent Reviewer
Linda Rothstein

Assistant Counsel
Robert Centa
Tina Lie
Jodi Martin

Communications
Peter Rehak

Editors
Dan Liebman
Mary McDougall Maude
Rosemary Shipton

Executive Coordinator
Joanna Arvanitis

Translation
Larrass Translations
Suzanne Labbé

Production Services
Laura Brady