A retrospective study of 191 epilepsy monitoring unit admissions at Mayo Clinic Arizona found that cannabis users had shorter hospital stays and higher rates of seizure capture during monitoring, but they also carried substantially more psychiatric diagnoses and trauma histories. These confounding differences make it impossible to attribute the monitoring outcomes to cannabis use itself, and the findings should be considered hypothesis-generating rather than practice-changing.

Epilepsy monitoring unit admissions represent a resource-intensive, high-stakes step in seizure diagnosis and surgical planning. As cannabis legalization expands across the United States, clinicians are encountering a growing proportion of EMU patients who use cannabis regularly, yet almost nothing is known about how cannabis use intersects with monitoring outcomes, discharge timing, or seizure capture success. Understanding these dynamics is critical for anticipating clinical complexity, planning admission protocols, and counseling patients about what to expect during hospitalization when their usual cannabis access is removed.

Epilepsy monitoring units are the primary setting for prolonged video-EEG recording, where clinicians aim to capture a patient’s typical seizure events to classify spell type, localize seizure foci, and guide treatment decisions including possible surgery. Admission duration is influenced by how quickly diagnostic events occur, and event capture rates are a key performance metric. This retrospective study from Mayo Clinic in Phoenix, Arizona, examined all 191 eligible EMU admissions in 2023 to determine whether cannabis use, identified by self-report or positive urine drug screen, was associated with differences in length of stay, event capture, or clinical profiles. Because cannabis is prohibited during hospitalization, the investigators hypothesized that forced abstinence might impair monitoring outcomes for regular users.

Cannabis users (61 patients, or 31.9% of the cohort) had a mean EMU length of stay approximately 0.9 days shorter than non-users, and among patients admitted specifically for spell classification, event capturability was 18.1% higher in the cannabis group. However, cannabis users also demonstrated substantially higher rates of major depressive disorder (an 18.9 percentage point difference), generalized anxiety disorder (22.1 percentage point difference), and histories of physical abuse (12.6%), sexual abuse (11.1%), and mental abuse (10.2%). These profound psychiatric and psychosocial differences between groups are likely independent predictors of both EMU behavior and seizure phenotype, and the available text does not confirm that multivariable adjustment for these confounders was performed. The authors acknowledge that the findings are preliminary and call for larger, prospective, multicenter studies.

This study asks an important question that clinicians managing epilepsy patients are quietly wrestling with every day: does cannabis use change what happens during a hospital monitoring admission? The finding that cannabis users had shorter stays and better event capture is genuinely interesting, but I cannot responsibly point to cannabis as the reason. When one group carries dramatically more depression, anxiety, and trauma, the observed differences in monitoring outcomes could reflect countless pathways that have nothing to do with the plant itself. The psychiatric burden alone could alter seizure threshold, stress reactivity, and even the decision to request early discharge.

In my practice, I see many patients who use cannabis and live with epilepsy, and I take a detailed history of both. Before any planned EMU admission, I discuss what forced abstinence might feel like, address the psychiatric comorbidities that this study rightly highlights as prevalent in this population, and coordinate closely with the neurology team. I do not promise patients that their cannabis use will help or hinder the monitoring process, because we simply do not have reliable data to support either claim. What I do emphasize is that their psychiatric health during hospitalization deserves just as much clinical attention as the EEG tracings.

This study sits very early in the research arc connecting cannabis use to epilepsy monitoring outcomes. While a substantial body of evidence supports the efficacy of pharmaceutical cannabidiol (Epidiolex) for specific epilepsy syndromes, the question of how recreational or medical cannabis use by epilepsy patients affects the diagnostic monitoring process is largely uncharted territory. The finding of a 31.9% cannabis use prevalence rate among EMU admissions is itself notable and consistent with rising population-level cannabis use, underscoring that this is a clinical reality epileptologists and hospitalists must increasingly navigate regardless of the evidence base for outcomes.

Clinicians should note that forced cannabis abstinence during hospitalization introduces a pharmacological variable that this study design cannot adequately isolate. Cannabinoid withdrawal can produce irritability, sleep disruption, and autonomic changes, all of which might plausibly alter seizure threshold or patient tolerance for prolonged monitoring. The elevated psychiatric comorbidity burden in the cannabis group also has direct implications for medication management during admission, as interactions between antiepileptic drugs, psychiatric medications, and residual cannabinoid effects warrant careful pharmacovigilance. The most actionable takeaway from this study is not about cannabis per se, but about ensuring that EMU admission protocols include systematic screening for psychiatric comorbidities and trauma histories in cannabis-using patients, who appear to carry a disproportionately high burden of both.