Background The impact of medical cannabis on healthcare utilisation between 2014 and 2017 in Ontario, Canada. With cannabis legalisation in Canada and some states in the USA, high-quality longitudinal cohort research studies are of urgent need to assess the impact of cannabis use on healthcare utilisation.

Methods A matched cohort study of 9925 medical cannabis authorised adult patients (inhaled (smoked or vaporised) or orally consumed (oils)) at specialised cannabis clinics, and inclusion of 17 732 controls (not authorised) between 24 April 2014 and 31 March 2017 from Ontario, Canada. Interrupted time series and multivariate Poisson regression analyses were conducted. Medical cannabis impact on healthcare utilisation was measured over 6 months: all-cause physician visits, all-cause hospitalisation, ambulatory care sensitive conditions (ACSC)-related hospitalisations, all-cause emergency department (ED) visits and ACSC-related ED visits.

Results For medical cannabis patients compared with controls, there was an initial (within the first month) increase in physician visits (additional 4330 visits per 10 000 patients). However, a numerical reduction was noted over the 6-month follow-up, and no statistical difference was observed (p=0.126). Likewise, in hospitalisations and ACSC ED visits, there was an initial increase (44 per 10 000 people, p<0.05) but no statistical difference after follow-up (p=0.34). Conversely, no initial increase in all-cause ED visits was observed with a slight decrease (19 visits per 10 000 patients, p=0.014) in follow-up.

Conclusions An initial increase (within first month) in healthcare utilisation may be expected among medical cannabis users that appears to wane over time. Proactive follow-up of patients using medical cannabis is warranted to minimise initial risks to patients and actively assess potential benefits/harms of ongoing use.