Abstract
Aims
To identify variation in identification of medical consumers using alternative self-reported measures and assess whether differences in these rates exist across jurisdictions with different medical policy approaches using evidence from an international study on cannabis use.
Design
Secondary analysis of wave 4 (2021) of the International Cannabis Policy Study (ICPS) cross-sectional survey.
Setting
United States, Canada and Australia.
Participants
16 951 (USA 10 472; CAN 5935; AUS 544) respondents who completed the survey and reported past year cannabis use across the three jurisdictions.
Measurements
Four different medical cannabis use measures were available, and rates of each were estimated using logistic regression methods that adjusted for age, gender, education and ethnicity. Medical cannabis use measures included potentially authorized use (i.e. involving a licensed health professional recommendation, authorization or prescription), pharmaceutical use (i.e. involving a pharmaceutical-grade product), therapeutic use (i.e. to manage physical or mental health conditions) and self-identified medical cannabis use. Country-specific differences were compared and discussed in light of measure and differing cannabis policies.
Findings
In wave 4 of the ICPS, 34.0% reported any past year cannabis use, but rates of medical use differed significantly according to the specific question. Far more individuals reported therapeutic use in the past year across all countries [77.3%; 95% confidence interval (CI) = 76.4%–78.2%] than any other measure of medical use. While just over one quarter (28.2%; 95% CI = 27.3%–29.2%) self-identified as a medical user, fewer reported being potentially authorized (22.8%; 95% CI = 22.0%–23.7%) or having a pharmaceutical prescription from a medical professional (12.3%; 95% CI = 11.6%–13.0%). Australians (27.2%; 95% CI = 23.0%–31.4%) and Americans (25.9%; 95% CI = 24.6%–27.2%) were more likely to report potentially authorized use than Canadians (17.3%; 95% CI = 16.1%–18.4%), but only Australians (27.4%; 95% CI = 23.6%–31.2%) reported high levels of prior use of a pharmaceutical-grade cannabinoid.
Conclusions
In the International Cannabis Policy Study, the proportion of respondents (adjusted for demographic factors) who reported medical use varied depending on the measures used within and between countries.
RATIONALE
Global interest in the medical use of cannabis has grown significantly in recent years despite limited supportive evidence from randomized control trials (RCTs). Studies of medical use using real-world data in naturalistic settings have struggled to address differences in the products used [1, 2], variability in the cannabinoids present in similarly named products [3, 4], differences in clinical treatment guidance [5] and differences among jurisdictions in which medical conditions qualify for cannabis use [1, 6, 7]. Surveys of the prevalence of medical cannabis use are affected by each of these factors and an additional one: whether the patient considers their use to be medical.
This study uses data from the International Cannabis Policy Study (ICPS) [8] to investigate how the specific questions posed about medical use can influence response rates in nationally representative samples of adults from the USA, Canada and Australia, three countries with distinctly different medical markets.
In the USA, 76% of states have legalized medical cannabis use as of June 2024, and more than half of those have also legalized non-medical ‘adult use’ [9]. USA federal law, however, continues to prohibit the use of cannabis for either purpose. In Canada, legal access to medical cannabis was established in 1999, although the number of individuals seeking medical authorizations has decreased substantially since the federal legalization of non-medical cannabis use in 2018 [10]. Medicinal cannabis use was legalized federally in Australia in 2016, although non-medical cannabis use remains illegal and states differ in the penalties imposed for non-medical use [11].
This study assessed how survey participants responded to different questions related to medical cannabis use within each country as well as across countries. To take account of national differences in terminology (e.g. medical marijuana, medicinal cannabis, medical cannabis products, etc.), the term ‘medical cannabis’ will be used throughout this article to refer to the medical use of a product that contains cannabinoids. The term ‘marijuana’ will be used only in reference to specific questions in the ICPS survey that used this term.
DATA, MEASURES & METHODS
Study design and data
Cross-sectional data from Wave 4 of the 2021 ICPS were collected via self-completed web-based surveys administered from September through October 2021. The target population are individuals aged 16–64 years residing in households. Samples were obtained via the Nielsen Consumer Insights Global Panel and partner panels. For the ICPS survey, Nielsen draws stratified random samples from its online panels, using quotas based on age (16–64 years) and state/province of residence. These respondents receive remuneration to participate in ICPS in accordance with their panel’s incentive structure upon survey completion. Post-stratification sample weightings are constructed based on age, sex and education to generate nationally representative samples for each country. Surveys are conducted in English (in the USA, parts of Canada and Australia) or French (in parts of Canada). More than 49 950 individuals responded to the ICPS Wave 4 survey across the three countries, of which two-thirds (64.4%) responded to questions on past-year cannabis use.
The study received ethics clearance from the University of Waterloo Research Ethics Committee (ORE#31330). For further details on study methods, see the technical reports and methodology article [8, 12].
Medical cannabis use measures
The survey contains multiple items pertaining to medical use, allowing us to construct four different measures of medical cannabis use (for the exact questions posed, see Figure 1). The first two, asked of all respondents, focused on medical use through a healthcare provider. Potentially authorized users are defined as having ever asked for or received a recommendation, authorization or prescription to use medical marijuana from a licensed health professional. Pharmaceutical Rx users are defined as those who report being prescribed a pharmaceutical-grade cannabis medication in the past 12 months, including dronabinol, Epidiolex®, nabilone, nabiximols or a ‘prescribed cannabis-based medicinal product’. The third measure, therapeutic use, was only asked of respondents who report ever trying cannabis and was intended to capture use to manage symptoms from specified mental or physical health conditions. Finally, self-identified medical use was based on a question posed of past-year cannabis users who were asked whether they self-identified as a medical cannabis user. For this last measure, the question was randomized in the survey so that 50% of those surveyed were asked if they only viewed themselves as taking cannabis for medical purposes, whereas the other 50% were asked if they consider themselves a ‘medical marijuana user’. Responses of ‘do not know’ were coded as non-medical use and ‘refused to answer’ were coded as missing for our analyses. These four medical use measures are not mutually exclusive (e.g. a self-identified individual may also be an individual reporting therapeutic use).
Read full paper at
