Only 18% of prescribed cannabinoid patients in Manitoba were still using them after one year | Research

The takeaway

In a population-based study of 5,452 new cannabinoid prescription users in Manitoba, only 18% continued use at one year, with median duration of just 31 days and significant variation by diagnosis and medication type.

Prescribers considering cannabinoid medications, health policy researchers, and patients starting prescribed cannabinoids.

18% still using at 1 year

What the researchers found

Among 5,452 new prescribed cannabinoid users, 18.1% continued use at 1 year. Median duration was 31 days. Nabilone had the longest median use (33 days) and nabiximols the shortest (20 days). Fibromyalgia, osteoarthritis, and substance use disorder were associated with longer use, while cancer was associated with much shorter use.

Why it matters

Real-world persistence data reveals that most prescribed cannabinoid patients stop relatively quickly. The reasons differ by condition: cancer patients may stop due to disease progression or death, while chronic pain patients may persist longer because their conditions are ongoing.

The numbers in context

5,452 new users studied. 18.1% (95% CI 17.08-19.12) still using at 1 year. Median duration: 31 days (IQR 25-193). Nabilone median: 33 days. Nabiximols median: 20 days. Cancer patients had HR 2.73 for discontinuation (95% CI 2.02-3.67). Substance use disorder had HR 0.85 for discontinuation.

How the study worked

Retrospective population-based cohort study using administrative health data from the Manitoba Population Research Data Repository. Included all new cannabinoid prescription users from April 2004 to April 2016, followed for 1 year.

What this study cannot tell us

Administrative data cannot capture reasons for discontinuation (inefficacy, side effects, symptom resolution, or death). Cannabis obtained outside the medical system was not tracked. The study period predated Canadian recreational legalization.

How to read the evidence

Rated moderate because this is a large population-based cohort study with over 5,000 subjects and 12 years of data, though it lacks information on reasons for discontinuation.

When this study was published

Published in 2019, covering prescriptions from 2004-2016. Canadian recreational cannabis was legalized in 2018, which may have changed prescribing patterns.

The bigger picture

Prescription persistence is a proxy for real-world effectiveness and tolerability. The finding that most patients stop within a month suggests either the medications are not meeting expectations, side effects are problematic, or the conditions they are treating are self-limited.

Questions still open

Why do most patients stop so quickly? Does the low persistence reflect poor efficacy, intolerable side effects, or access to non-prescription cannabis? How has recreational legalization affected medical cannabis prescription patterns?

Common questions

How long do people typically use prescribed cannabinoids?

In this study, the median was only 31 days. Only about 1 in 5 patients were still using them after a year.

Which conditions led to longer use?

Fibromyalgia, osteoarthritis, and substance use disorder were associated with longer persistence. Cancer patients tended to stop much sooner (2.7 times more likely to discontinue).

Does this mean cannabinoid medications do not work?

Not necessarily. The study could not determine why patients stopped. Possible reasons include side effects, symptom resolution, switching to non-prescription cannabis, or disease progression.

Read the original research

Persistence of use of prescribed cannabinoid medicines in Manitoba, Canada: a population-based cohort study.

Addiction (Abingdon, England), 114(10), 1791-1799

View on PubMed ↗Publisher’s paper ↗

Citation

Alkabbani, Wajd; Marrie, Ruth Ann; Bugden, Shawn; Alessi-Severini, Silvia; Bolton, James M; Daeninck, Paul; Leong, Christine. (2019). Persistence of use of prescribed cannabinoid medicines in Manitoba, Canada: a population-based cohort study.. Addiction (Abingdon, England), 114(10), 1791-1799. https://doi.org/10.1111/add.14719