Cerium Journal Club – 1st Tuesday of Every Month!
Can Brain Imaging Detect THC Impairment Better Than Field Sobriety Tests?
Time: July 7th, Noon EST
Location: Virtual Meeting
https://us02web.zoom.us/j/83722922366?pwd=QjZ3V0xNWHZ0OVZ0b3BpbWd3STRqUT09
Event Description
This month’s Cerium Journal Club will examine a new randomized clinical trial published in JAMA Network Openexploring whether portable brain imaging can detect THC-related impairment more accurately than conventional field sobriety testing.
Cannabis impairment remains one of the most difficult issues in clinical practice, public safety, forensic interpretation, and workplace policy. Unlike alcohol, THC levels in blood, saliva, or urine do not reliably indicate whether someone is functionally impaired. Regular cannabis use, tolerance, route of administration, individual metabolism, and lingering metabolites all complicate interpretation.
This study tested whether functional near-infrared spectroscopy, or fNIRS, could identify a neural signature of THC-induced impairment by measuring prefrontal cortex activity. Participants received oral synthetic THC or placebo in a double-blind, randomized crossover design. Machine learning models trained on fNIRS data were then compared with field sobriety testing for identifying clinically determined impairment.
The findings are provocative: resting-state fNIRS classified THC-induced impairment with higher accuracy and a much lower false-positive rate than field sobriety testing. The paper raises important questions about what we mean by “impairment,” how we define ground truth, and whether objective-looking technologies are ready for roadside, workplace, clinical, or legal use.
Featured Article
Berchansky M, Evins AE, Evohr B, Himmelsbach Z, Pachas GN, Karunakaran KD, Laufer Goldshtein B, Ozana N, Gilman JM.
Detection of Δ9-Tetrahydrocannabinol Impairment Using Resting-State Functional Near-Infrared Spectroscopy: A Randomized Clinical Trial.
JAMA Network Open. 2026;9(1).
Key takeaway: A 6-minute resting-state fNIRS assessment identified THC-related impairment with greater accuracy and fewer false positives than expanded field sobriety testing, suggesting that portable neuroimaging may offer a promising but still early-stage approach to cannabis impairment detection.
Discussion Questions
What does this study tell us about the limits of THC concentration as a marker of impairment?
Is fNIRS measuring impairment, intoxication, compensation, tolerance, or some mixture of all four?
How convincing is the study’s definition of “clinical impairment” as the ground truth?
What are the strengths and weaknesses of using machine learning models for impairment classification?
Could portable neuroimaging ever be practical in roadside, workplace, clinical, or research settings?
What ethical, legal, and public health concerns arise when brain-based tools are proposed for impairment detection?
How should clinicians, regulators, employers, and courts interpret emerging technologies that appear more objective than behavioral assessments?
Does this kind of technology reduce bias, or could it introduce new forms of bias through study design, training data, interpretation, or deployment?
Speakers
Phil Molloy, MD – Clinical education expert
Teresa Simon, MPH – Public health epidemiologist
Len Kamen, MD – Clinical pain specialist
Jahan Marcu, PhD – Cannabis researcher
Questions?
Email: PRCtrials.info@gmail.com
This journal club is supported by PRC+.
PRC+ has education modules and resources available for healthcare professionals:
https://www.prc-trials-plus.com/education