Presenting at AAFS 2026: Errors in Toxicology Testing and the Need for Full Discovery
How discovery violations, systematic errors, programming defects, and deliberate misconduct undermined justice across multiple jurisdictions
I’m pleased to announce that I’ll be presenting at the 78th Annual Scientific Conference of the American Academy of Forensic Sciences (AAFS) in New Orleans this February. My presentation, “Errors in Toxicology Testing and the Need for Full Discovery,” will take place on Friday, February 13, 2026, from 9:00 to 9:45 AM.
This presentation is based on a paper published in Forensic Science International: Synergy, co-authored with defense attorney Charles Ramsay, that documents notable toxicology errors we’ve collected over a combined 48 years in the field. The research reveals systemic vulnerabilities that have affected thousands of criminal cases across multiple jurisdictions.
The Scope of the Problem
While toxicology is often viewed as one of the more objective forensic disciplines due to its foundation in analytical chemistry and quantitative measurements, our research demonstrates that it remains vulnerable to errors ranging from technical failures to deliberate misconduct.
The cases we examined fall into distinct categories: traceability errors, calibration failures, discovery violations, maintenance problems, source code defects, fraud, interfering substances, reporting errors, laboratory contamination, and chain of custody breaches. What emerged were troubling patterns that reveal systemic weaknesses in how forensic toxicology is practiced and overseen.
Patterns That Demand Attention
Errors Persisted for Years
Many of the errors we documented went undetected for extended periods. The District of Columbia miscalibrated its breath alcohol analyzers 20-40% too high for 14 years before a new employee discovered the problem. Maryland’s laboratory used scientifically invalid single-point calibration from 2011 onward, somehow passing accreditation visits in 2015 and 2019 before being cited for non-conformity in 2021.
External Discovery, Not Internal Controls
Perhaps most concerning is that errors were typically discovered by external sources rather than internal quality assurance systems. Defense attorneys, whistleblowers, and new employees, not established laboratory personnel or accreditation bodies, identified the problems. This pattern suggests that self-monitoring systems are insufficient and that laboratories can develop insular cultures resistant to critical evaluation.
Systematic Discovery Violations
The Massachusetts Office of Alcohol Testing maintained a “longstanding but unwritten policy” of withholding worksheets documenting failed calibrations. When this came to light, approximately 27,000 drunk driving cases became eligible for new trials or withdrawn guilty pleas due to what the Massachusetts Supreme Judicial Court called it “egregious government misconduct” and “a disturbing pattern of intentionally withholding exculpatory evidence year after year.”
This wasn’t an isolated incident. In Arizona, forensic scientist Gregory Ohlson, with 40 years of experience, was forced into early retirement after testifying that full disclosure of batch data from blood alcohol analysis could reveal accuracy problems. In Idaho, courts initially excluded evidence of calibration failures, requiring defendants to have expert witnesses even to cross-examine state witnesses about the issues.
Retaliation Against Whistleblowers
When problems were identified internally, those who raised concerns often faced retaliation. Gregory Ohlson’s supervisors told him he needed to change his testimony to “align” with other laboratory scientists. Vermont toxicologists Darcy Richardson and Amanda Bolduc documented that a technician was manipulating breath alcohol analyzer certification results, yet the laboratory initially dismissed their allegations.
Real-World Consequences
The impact of these errors extends far beyond laboratory walls. In the Randox Testing Services scandal in the UK, fraudulent toxicology testing affected over 10,000 criminal cases across 42 police forces between 2013 and 2017, spanning road traffic offenses, sexual offenses, violent crimes, and unexplained deaths.
Clinical laboratory errors have torn families apart. Multiple mothers, Eileen Bower, Elizabeth Mort, and Elizabeth Eden, had their newborn infants removed from their custody based on false-positive opiate screens triggered by consuming poppy seeds. These separations occurred because laboratories used inappropriately low cutoff thresholds (300 ng/mL versus the federal workplace standard of 2000 ng/mL) and failed to conduct confirmatory testing.
The Motherisk Drug Testing Laboratory scandal in Canada led to over 35,000 flawed hair drug tests between the late 1990s and 2015, resulting in wrongful removal of children from families. The laboratory operated without forensic accreditation and misused screening tests as confirmatory evidence, a practice described by an independent review as something “no forensic toxicology laboratory in the world” would do.
The Discovery Dilemma
These cases reveal a fundamental tension in forensic science between supporting prosecution and maintaining scientific objectivity. The prevalence of discovery violations suggests many forensic laboratories either misunderstand or deliberately circumvent their legal obligations under Brady v. Maryland, treating themselves as extensions of law enforcement rather than as scientific institutions bound by ethical and legal duties to disclose exculpatory evidence.
Breath alcohol testing presents unique challenges. Unlike blood or urine samples that can be retained and retested, breath samples are destroyed during analysis. This ephemeral nature makes the disclosure of digital data and calibration records essential. Without access to complete analytical records, including expirograms, flow rates, and real-time data, independent experts cannot properly verify whether tests were properly conducted.
Source Code: The Invisible Problem
Software defects present particularly insidious challenges because they can affect thousands of tests while remaining invisible to operators. We documented programming errors in breath alcohol analyzers across multiple jurisdictions:
Massachusetts: The Dräger Alcotest 9510 was programmed with tolerance ranges that didn’t match state requirements, requiring manual verification and introducing human error
Minnesota: The Intoxilyzer 5000EN had a programming error, causing valid samples to be rejected; the laboratory received a patch in 2007 but didn’t install it to avoid drawing attention
New Jersey: Independent analysis of the Dräger Alcotest 7110 source code revealed thousands of programming errors using outdated methodologies
Washington: Software issues caused inconsistent, invalid sample message rates, fluctuating from 4.3% to 40.5% across different versions
Many jurisdictions resist source code disclosure, citing trade secret protections. Yet without independent verification, these defects remain hidden.
The Path Forward
Based on our analysis, several reforms are essential:
Transparency Through Online Discovery Portals: Laboratories should provide online access to validation studies, standard operating procedures, maintenance records, quality assurance reports, and incident documentation. Massachusetts State Police implemented such a portal following their breath testing scandal; other jurisdictions should follow.
Mandatory Retention and Disclosure of Digital Data: All calibration records, chromatograms, quality assurance charts, instrument maintenance logs, and metadata should be preserved and made available. For breath alcohol testing specifically, digital expirograms, flow rates, and real-time monitoring data must be retained.
Accessible Data Formats: Laboratories should provide data in formats that facilitate independent analysis, such as tabular data in spreadsheet formats rather than PDFs.
Source Code Access: Source code for analytical instruments should be available for independent review under appropriate protective orders, or manufacturers should create open-source repositories for public review.
Laboratory Independence: Forensic laboratories should strive for independence from law enforcement agencies, as recommended by the National Academy of Sciences.
Rigorous Accreditation: Accreditation bodies must conduct more thorough assessments. Maryland’s laboratory passed multiple audits while using invalid single-point calibration, demonstrating current oversight inadequacies.
Whistleblower Protections: Legal protections must be established for forensic scientists who report problems or violations of discovery. The retaliation faced by those who speak up creates a chilling effect that perpetuates errors.
Regular Third-Party Audits: Beyond accreditation requirements, independent experts should conduct regular audits of laboratories. Many documented errors persisted for years and were only discovered through aggressive discovery efforts.
Looking Ahead
The forensic science community has made tremendous progress in improving laboratory practices and standards. But there’s more work to be done, particularly regarding transparency, accountability, and protecting those who identify problems.
My hope is that this presentation will contribute to an ongoing conversation about strengthening forensic toxicology. By examining past errors honestly, we can develop policies that prevent similar mistakes and ensure the results society relies upon are truly trustworthy.
These aren’t just technical or procedural issues; they’re fundamental to justice. When toxicology results influence life-altering decisions in criminal prosecutions, child custody matters, and other legal proceedings, accuracy and transparency become moral imperatives.
If you’re attending AAFS 2026 in New Orleans, I’d welcome the opportunity to discuss these findings and potential solutions. And if you’ve experienced similar issues or have insights to share, please reach out.
The 78th Annual AAFS Scientific Conference will take place February 9-14, 2026, in New Orleans, Louisiana.
The full paper, “Errors in toxicology testing and the need for full discovery,” is published in Forensic Science International: Synergy and available as open access.