Adjusted Injury-Severity Findings Differed From Crude THC Trauma Comparisons
THC positivity alone was not an independent predictor of higher injury-severity categories in the study’s adjusted models, despite differences in simpler comparisons. The adjusted null needs to remain visible alongside the crude findings.
The models compared severity categories
Borst and colleagues analyzed toxicology-screened crash patients from five trauma centers. Their regression compared higher Injury Severity Score categories with scores of one through eight, using the toxicology-negative group as the reference. Adjustment included demographic, year, mechanism and institution-related variables.
The toxicology categories did not share one result
The paper reports that THC alone did not independently predict the higher severity categories after adjustment. Alcohol-positive, combined THC/alcohol and stimulant-related categories had increased odds in the reported models. Their findings should not be relabeled as the adjusted effect of THC alone.
Adjustment changes the question, not the observed injuries
The result does not show that THC protected patients or that cannabis cannot contribute to a collision. It estimates a conditional association among selected injured patients under the model’s assumptions. Residual confounding and selection remain possible. An unadjusted association also cannot be treated as an independent drug effect simply because it is significant. A sound interpretation identifies the outcome category, exposure group and adjustment set, then distinguishes injury severity among people already injured from the separate probability of having a crash. The study supplies neither a population driving-risk multiplier nor an individual impairment determination.
Related analysis: injury severity and population crash risk.
Research cited
Borst and colleagues. Driving under the influence: a multi-center evaluation of vehicular crashes in the era of cannabis legalization. Trauma Surgery & Acute Care Open (2021). DOI: 10.1136/tsaco-2021-000736.
