This population-based cohort study showed important characteristics about adult patients who were medically authorized for cannabis in ON and AB, Canada. Overall, the demographics of the adult population in ON and AB were comparable for both mean age and sex at the index date of authorization. The primary reasons for cannabis authorization were similar between ON and AB with pain, mental health, and sleep problems as the most common.

For the other 70% of patients in ON and AB, they were categorized into more than one high-level condition (pain, mental health, autoimmune, cancer, sleep problems, neurological, gastrointestinal, other, or uncategorized) for cannabis authorization, suggesting these patients have multiple comorbidities which are not currently being adequately addressed with more traditional medical therapies. This also may suggest that the nature of therapeutic applications of cannabis is multifaceted. For example, a patient experiencing cancer pain may also suffer from sleep problems. Hence, the multi-symptom [16] nature of patients who seek cannabis may make identification of the efficacy of medical cannabis, per symptom, difficult. From a quality of life standpoint, medical cannabis may be viewed as effective for one symptom (ex: sleep)—however, if improvement of pain or cancer were also included as outcomes, then cannabis may appear ineffective. These outcomes may be a contributing factor to the mixed results [2] on cannabis efficacy in the literature.

Previous research studies have reported a level of substantial evidence for medical cannabis’ use in the alleviation of pain [5, 27, 29]. Both Canadian clinicians [5] and the NASEM [30] have concurred that medical cannabis use may be effective in the management of chronic pain. In addition, they also suggest medical cannabis may be effective for chemotherapy-induced vomiting/nausea, and multiple sclerosis spasticity symptoms [5, 30]. Further, previous research has shown that there is an inherent psychological [31] element to medical cannabis use, in particular for relieving anxiety [4, 14] and depression [18, 19]. A significant portion of mental health outcome studies on medical cannabis also included its frequent utilization for both sleep problems [21, 22] and post-traumatic stress disorder [13, 15]. Otherwise, clinician recommendations emphasize limiting medical cannabis use for any other ailment (as a first-choice treatment plan) [5, 23].

Interestingly, this study shows that AB and ON physicians are prescribing medical cannabis for over 50 listed conditions (Additional file 1: Table S1). At least 25% of patients in this study were using medical cannabis for reasons not currently supported by evidence. Moreover, 25% is likely a very conservative estimate as it assumed that all reported reasons met the specific criteria which has the support of evidence (i.e., all cancer pain was refractory cancer pain, all patients stating diabetes and pain had neuropathic diabetic pain, all sleep problems were short term, etc.). This observation aligns with the systematic review by Lim et al. [12]—that shows that a significant portion of medical cannabis is used for a diverse range of specific health conditions or disorders. Furthermore, the systematic review showed little to no evidence supporting medical cannabis use on the treatment of these other health conditions. Likewise, NASEM [23] has also reported that there is limited to insufficient evidence for medical cannabis use on any health conditions other than chronic pain. This is also supported by Canadian evidence [5] on avoiding cannabis as a first line or second line of treatment. The guidelines, however, do make an exception for a small consideration of neuropathic pain—but this is emphasized as a weak recommendation.

Although limited evidence existed within reviews and guidelines for many of the conditions, that is not to suggest patients may not be experiencing benefit from medical cannabis for their symptoms of the condition but that the scientific evidence to support the effectiveness and association of medical cannabis with the health condition has not been established; or in many cases, research has not been conducted at all. However, it could also speak to potential lack of education available in regard to medical cannabis among physicians as recent studies have reported that Canadian physicians and physicians-in-training desire more education and knowledge in regard to the protocol and treatment plans for medical cannabis therapy [32]. Indeed, one of the critical gaps in their training was reported to be understanding how to implement effective medical cannabis treatment plans that accounted for all risks and benefits of medical cannabis [33]. Conversely, other physicians [34] have also reported hesitancy or complete reluctance to authorize medical cannabis as a result of this lack of guidance and education. Thus, our study provides ongoing surveillance for understanding physician prescribing behavior and potential educational needs on the evidence of medical cannabis for these certain types of health conditions.

It is important to consider that, although the use of medical cannabis for certain conditions may not be supported by current reviews or guidelines, the use of cannabis through the medical system (i.e., prescribed by physicians or nurse practitioners), as opposed to self-medication through non-medical retail sales which are now legal in Canada, may have important implications for harm reduction. Indeed, inclusion of physicians and other health professionals that have a full understanding of the patients comorbidities and other medication use would be expected to identify and reduce potential adverse effects of cannabis as patients can be actively monitored and assessed by qualified professionals. Moreover, if medical cannabis is felt to improve symptoms of a condition, there may be potential additional harm reductions with its use. For example, pain was one of the highest conditions noted where medical cannabis is being used in this cohort of patients. We have previously shown that cannabis may have important impacts to help reduce the use of opioids for chronic pain at the population level [35]. Thus, medical cannabis authorization could be a potential strategy to minimize harm associated with opioid use.

The major strength of this study is that it is currently Canada’s largest study on medical cannabis authorized adult patients. Therefore, the data from this study provide current and critical information about Canadian patients seeking cannabis as a form of therapy for various health reasons. The findings from this study contribute to the knowledge base on population-level use beyond its non-medical uses. Our analyses further provide evidence on important subgroups of patients who are currently using medical cannabis for treatment of over 200 conditions.

This study relied on previously published reviews or guidelines in the assessment of evidence. It is possible that other evidence exists for the use of medical cannabis within certain conditions that were not addressed in the previous reviews or guidelines and therefore would not have been included in our levels of evidence assessments. Another limitation to this study is that our data are restricted to only patients who have attended a cannabis medical clinic. It is also possible that responders to the questionnaires provide information overstating their health symptoms as a means to increase their likelihood of being authorized medical cannabis. We also did not appraise the level of evidence ourselves and used the appraisals from other studies (Additional file 1: Appendix A). The Canadian evidence [5] and the NASEM [23] guidelines often had differing opinions on the strength of evidence with the US guidelines saying substantive evidence and the Canadian evidence reporting weak evidence. When differences were found, the recommendations from the Canadian evidence recommendations were used. Finally, we have no method to determine whether cannabis was being considered as a first, second, or alterative line of therapy—in most cases it is more than plausible it was not used as first line.