Canada's Cannabis Surveillance Is Broken: Outdated Tools Miss the Harms Legalization Was Supposed to Track | Research

The takeaway

Seven years after legalization, Canada's national surveys still use outdated diagnostic frameworks, skip logic that excludes many users, and administrative data that underdiagnoses cannabis use disorder — making it nearly impossible to measure what legalization has actually done to public health.

Read this if you follow cannabis policy — Canada's experience shows that legalizing cannabis without adequate surveillance tools means you can't tell whether your policy is working.

What the researchers found

Canada legalized non-medical cannabis in 2018 as a public health initiative. The stated goal was to move cannabis from criminalization to regulation, with the expectation that regulated access would reduce harms. Seven years later, this commentary argues that Canada doesn't actually know whether that's happened — because the surveillance tools designed to measure cannabis harms are fundamentally inadequate.

The problems are specific and damning. National surveys still use DSM-IV diagnostic frameworks rather than the current DSM-5 criteria for cannabis use disorder. They employ skip logic that excludes individuals with lower levels of use from substance use questions — missing the very population where early problematic use might be detected. Key symptoms like loss of control, functional impairment, and withdrawal aren't assessed.

Administrative health records are equally problematic. Cannabis use disorder is underdiagnosed in clinical settings, inconsistently coded, and absent from many administrative databases. The primary mental health surveillance tool doesn't align with DSM-5 standards. High-risk populations — including youth, Indigenous communities, and people experiencing homelessness — are often excluded from survey samples entirely.

The result: CUD is systematically underdetected, underreported, and absent from the policy discussions it should be informing. Canada ran a massive public health experiment (legalization) without equipping itself to measure the outcomes.

Why it matters

Other countries and jurisdictions are using Canada as a model for cannabis legalization. If Canada's own surveillance systems can't determine whether legalization has increased or decreased harms, the evidence base for global cannabis policy is undermined. This isn't an argument against legalization — it's an argument that any public health intervention this large requires adequate measurement tools, and Canada's are currently not up to the job.

The numbers in context

Canada legalized non-medical cannabis in 2018. Major national surveys still use DSM-IV (superseded by DSM-5 in 2013). Skip logic in surveys excludes individuals with lower use levels. CUD is described as rarely measured, often misclassified, and largely absent from policy discussions. High-risk populations are frequently excluded from survey samples.

How the study worked

Narrative review and commentary examining Canada's current cannabis surveillance infrastructure, including national surveys, diagnostic frameworks, administrative health records, and case-finding tools. Assessed alignment with current diagnostic standards (DSM-5) and capacity to detect cannabis use disorder and related harms.

Who was studied

Not enough information provided.

What this study cannot tell us

This is a commentary/perspective piece rather than original research. Focused specifically on Canada — other jurisdictions may have better or worse surveillance. The author's argument that surveillance is inadequate is well-supported by examples, but others might argue that existing data captures enough for policy-making. The piece doesn't quantify how much CUD is being missed, only argues that current tools are insufficient to capture it accurately.

How to read the evidence

Commentary/narrative review examining surveillance methodology. Well-argued with specific examples but doesn't present original data. The critique of measurement tools is methodologically sound; the conclusions about underdetection are logical but not empirically quantified.

When this study was published

Published in 2025. Canada's surveillance tools may be updated in response to critiques like this one.

The bigger picture

This is the surveillance and policy counterpart to the clinical and biological studies in the database. Studies like RTHC-00115 (potency trends) and RTHC-00103 (cardiovascular meta-analysis) generate data about cannabis harms, but if national surveillance systems can't detect those harms in the population, the research findings never translate into effective public health responses. For other countries considering legalization, this is a cautionary tale about building measurement infrastructure before making policy changes.

Replication

Not stated in abstract.

Funding

Not reported in abstract.

Conflicts of interest

Not reported in abstract.

Questions still open

Would updating Canada's surveys to DSM-5 criteria and removing skip logic for lower-level users reveal significantly higher CUD rates? Can other countries that are legalizing cannabis learn from Canada's surveillance gaps and build better systems from the start? Would routine CUD screening in primary care settings improve detection? Is the underdetection of CUD affecting treatment access — are people not getting help because their condition isn't being identified?

Read the original research

Outdated tools, underestimated harm: Modernizing cannabis surveillance in a post-legalization era.

Addiction (Abingdon, England)

Addiction is a well-respected journal that publishes research on substance use and addiction.

View on PubMed ↗Publisher’s paper ↗

Citation

Bahji, Anees. (2025). Outdated tools, underestimated harm: Modernizing cannabis surveillance in a post-legalization era.. Addiction (Abingdon, England). https://doi.org/10.1111/add.70274