Cannabis use and anaesthetic agent dosing in patients undergoing surgery or interventional procedures: a single-centre retrospective cohort study.

BACKGROUND

There is limited evidence that habitual cannabis consumption increases anaesthetic requirements. We hypothesised that cannabis use affects anaesthetic dosing among patients undergoing surgery or interventional procedures.

METHODS

A total of 341 199 adult patients undergoing propofol sedation or volatile-based general anaesthesia at Beth Israel Deaconess Medical Center, Boston, USA, between 2008 and 2024 were included. The primary exposure was cannabis use, differentiated into medical or non-medical use (recreational use or cannabis use disorder). Primary outcomes were propofol dose (μg kg-1 min-1) for monitored anaesthesia care and mean end tidal age-adjusted minimum alveolar concentration (MAC) values for general anaesthesia. In multivariable analyses, we estimated the adjusted difference in mean primary anaesthetic agent dosing.

RESULTS

A total of 23 896 patients (7.0%) were identified as non-medical cannabis users, and 3193 (0.9%) held a certification for medical cannabinoids. In adjusted analyses, both non-medical and medical cannabis use were associated with increased propofol doses (non-medical, +6.85 μg kg-1 min-1, 95% CI 6.07-7.63, P<0.001; medical, +2.87 μg kg-1 min-1, 95% CI 0.88-4.86, P=0.005) and doses of volatile anaesthetics (non-medical, +0.01 MAC, 95% CI 0.01-0.02, P<0.001; medical, +0.02 MAC, 95% CI 0.01-0.03, P=0.002). Estimated effects were strongest among frequent users (daily recreational users, +8.41 μg kg-1 min-1 propofol, 95% CI 7.00-9.81, P<0.001). At comparable propofol induction doses, non-medical cannabis users experienced greater heart rate increase after tracheal intubation (+1.45 beats min-1, 95% CI 1.10-1.79, P<0.001).

CONCLUSIONS

Cannabis use was associated with modestly higher propofol and volatile anaesthetic doses. However, effect sizes were smaller than previously reported, and future studies should investigate the clinical relevance.