Adjustment Changed the Fentanyl Inference in a Marijuana–Endoscopy Cohort
A marijuana–endoscopy study reported a significant initial fentanyl association, but its adjusted high-fentanyl estimate did not reach statistical significance. Both analyses belong in the interpretation.
The participants were not randomized to marijuana use
Kosirog and colleagues prospectively studied 976 outpatients at one Veterans Affairs center. Marijuana history came from a voluntary survey, while sedation and other clinical information came from records.
Users differed from nonusers in age, body habitus, several comorbidities, smoking and vaping. Such differences make the adjusted analysis important rather than optional.
A positive point estimate was not a significant result
The initial fentanyl comparison had a reported p value of .003. In the multivariable high-fentanyl analysis, the odds ratio was 1.54, but its 95% confidence interval was 0.98–2.38. The authors reported that this comparison did not reach significance.
The adjusted midazolam finding had a different interval, and propofol was assessed in monitored-anesthesia cases. Those results should remain attached to their drug, sedation setting and outcome definition.
Avoid one label for every endpoint
A general statement about higher sedation use can conceal differences between continuous amounts, high-use categories and adjusted associations. Neither statistical adjustment nor a positive estimate establishes the biological mechanism.
The study supports examining cannabis history in the clinical context. It does not provide a universal medication multiplier or prove that cannabis tolerance caused each observed association.
Related: Interpreting cannabis tolerance claims.
Research cited
Kosirog J; Bouvette C; Pannu J; Gondal J; Madhoun M Gastrointestinal endoscopy · 2024 · 100(2):177 PMID 38215858 · DOI 10.1016/j.gie.2024.01.013. Marijuana and endoscopy: the effects of marijuana on sedation..
