A scoping review of 34 studies finds that while knowledge deficits dominate the medical cannabis literature, critical behavioural determinants such as professional identity, social norms, and environmental context remain largely unstudied. Fixing what clinicians know may not be enough to change how they practice if the broader forces shaping their behaviour are never addressed.

Medical cannabis is now legal in dozens of jurisdictions worldwide, yet patients routinely encounter clinicians who feel unprepared, uncomfortable, or unwilling to engage with it. The default assumption has been that this gap is a knowledge problem, solvable with more education. This review challenges that assumption at a structural level, arguing that the field’s nearly exclusive focus on what clinicians know has crowded out investigation of equally important forces: professional culture, peer influence, institutional environments, and the identity-level discomfort that no lecture can resolve.

Medical cannabis legalization has advanced rapidly in many jurisdictions, but clinical implementation has not kept pace. Healthcare professionals frequently report feeling underprepared to discuss, recommend, or prescribe cannabis-based therapies, raising the question of what exactly is holding them back. O’Rourke and colleagues set out to answer this question through a scoping review that applies the Theoretical Domains Framework, a validated behaviour change model integrating 14 domains, 128 constructs, and 33 psychological theories, as a systematic lens for examining 34 published studies spanning 2000 to May 2019 and encompassing 7,349 healthcare professionals across multiple countries.

The review’s central finding is a pronounced imbalance in the literature: all 34 studies addressed the knowledge domain, while beliefs about consequences (particularly concerns about misuse, dependence, and psychosis) appeared in 21 studies. By contrast, domains such as social and professional role identity, social influences, and environmental context were rarely examined despite their well-established theoretical importance for behaviour change. Most clinicians reported learning about medical cannabis informally, from patients, peers, and news media, rather than through structured continuing medical education. The authors acknowledge that as a scoping review, this work maps evidence breadth without assessing study quality or establishing causal relationships, and note that the May 2019 search cutoff means a substantial body of more recent post-legalization research is not captured. They conclude that future research and educational programs must explicitly incorporate social, environmental, and identity-based determinants of behaviour if consistent medical cannabis implementation is to be achieved.

Beyond Knowledge Gaps: Why Medical Cannabis Implementation Requires a Whole-Clinician Approach

If you asked a healthcare professional why they are uncertain about medical cannabis, they would probably tell you they do not know enough. They might be right, but they are almost certainly incomplete. This scoping review by O’Rourke and colleagues finds that the research community has been asking exactly the same narrow question, and in doing so, has systematically overlooked the professional identities, peer norms, and environmental realities that may matter just as much. What the review genuinely contributes is a structured demonstration of that blind spot: by coding 34 studies against the 14 domains of the Theoretical Domains Framework, the authors show that the entire literature concentrates on knowledge while leaving most other well-established behavioural determinants virtually untouched. That is a valuable finding. It transforms a vague sense that “education alone isn’t working” into a precise, domain-level map of where researchers have looked and where they have not. The analogy I keep returning to is this: telling a clinician more facts about cannabis to change their prescribing behaviour is like giving someone a map to change how they drive. Useful, certainly, but ineffective if the road signs, their passengers, and their GPS are all pointing elsewhere. Knowledge is necessary but plainly insufficient.

That said, I want to be honest about where the review’s inference gets ahead of its evidence. Identifying that a domain is underrepresented in the literature is not the same as demonstrating that the domain actually drives clinical behaviour. The TDF was applied post hoc to studies that were never designed with it in mind, which is a bit like trying to assess whether a kitchen is well equipped by reading restaurant reviews that never mentioned kitchen equipment. Absence of mention tells you about what reviewers focused on, not about what is actually in the kitchen. No inter-rater reliability data are reported for the domain coding, and the 51% US geographic skew and May 2019 search cutoff further constrain what we can confidently conclude. The review also does not distinguish between barriers to initiating a conversation about medical cannabis and barriers to actually prescribing it. These are different clinical behaviours with potentially different domain profiles. And it largely overlooks the patient side of the equation: in my clinical experience, patient advocacy and patient-clinician interaction dynamics are among the most potent forces driving how quickly a clinician’s comfort evolves.

What would I say to a patient who asks why their doctor seems uncomfortable discussing cannabis? I would say that your doctor’s hesitation probably is not just about not knowing enough. It is also about their training culture, their colleagues’ attitudes, and the regulatory environment they practice in. To a colleague, I would frame this review as a prompt: ask yourself whether your discomfort with medical cannabis is driven by genuine evidence concerns, by what your professional tribe has normalized, or by structural barriers in your practice environment, and then address all three rather than just reading more papers. In any rapidly legalizing or newly medicalized area, the first wave of research almost always focuses on what clinicians know. But the evidence base for sustainable behaviour change tells us we must study the whole clinician: their identity, their peers, their environment, and their beliefs, not just their knowledge scores.

For clinicians who have observed the disconnect between expanding medical cannabis legalization and the slow pace of clinical adoption, this review provides a theoretical framework for understanding why. The field is still in an early research arc where the dominant paradigm treats implementation failure as a knowledge problem. This review suggests that paradigm is incomplete and that the next generation of research needs to investigate professional identity, peer norms, and environmental context as independent predictors of clinical behaviour. The TDF provides a practical roadmap for where that research should begin.

From a pharmacological and safety standpoint, clinician concerns about misuse, dependence, and psychiatric harm documented across 21 of the 34 studies are not unfounded, though they may be disproportionate to the current evidence base for many patient populations. Drug interaction awareness, dosing uncertainty, and the absence of standardized formulations remain genuine clinical barriers that knowledge-focused education can partially address. The most actionable recommendation from this review is that medical education programs targeting cannabis should be designed with multi-domain frameworks in mind, incorporating not just pharmacological content but also opportunities for peer dialogue, identity reflection, and institutional support, rather than relying on information transfer alone.

This is a scoping review following PRISMA reporting guidelines and Cochrane methodology, applying the 14-domain Theoretical Domains Framework as a post hoc coding lens to 34 eligible publications. Scoping reviews map the breadth and focus of a literature base but do not formally assess study quality, pool quantitative estimates, or establish causal relationships. The single most important inference constraint is that domain frequency counts reflect the published literature’s focus, not the actual relative importance of different behavioural determinants in shaping real-world clinical practice.

This review builds on and extends an earlier systematic review by Gardiner and colleagues (2019), which similarly documented knowledge gaps and attitudinal barriers among healthcare professionals but did not apply a structured behaviour change framework. By using the TDF, O’Rourke et al. add a layer of analytical precision that prior reviews lacked, enabling the identification of specific domain-level gaps rather than broad thematic shortcomings. The review’s findings are consistent with the broader implementation science literature, which has repeatedly demonstrated across many clinical domains that knowledge-only interventions produce modest and often unsustained behaviour change when environmental, social, and identity factors are not simultaneously addressed.

The most consequential analytic choice was the application of the TDF as a post hoc coding framework to studies that were not designed with it in mind. If independent double-coding with formal inter-rater reliability assessment had been conducted, the domain frequency counts could have shifted meaningfully, particularly for ambiguous domains like social influences and environmental context, where the boundary between what a study “addressed” versus merely “mentioned” requires interpretive judgment. Additionally, extending the search window beyond May 2019 would likely have captured a substantial number of post-legalization studies that may partially address the identified gaps, potentially moderating the review’s strongest conclusions about domain-level neglect.

The most likely overinterpretation is reading this review as evidence that knowledge gaps are the primary barrier to medical cannabis implementation. In fact, the review demonstrates only that knowledge is the most studied domain, not that it is the most important determinant of behaviour. These are categorically different claims. The domain frequency counts reflect where researchers have directed their attention, which is shaped by convenience, funding priorities, and publication norms, not by empirical measurement of which factors actually drive prescribing decisions. Similarly, the review’s emphasis on professional identity as a theoretically important domain should not be read as evidence that identity barriers have been documented in the cannabis context; they have been hypothesized on the basis of broader psychological theory, not measured in the included studies.

This scoping review contributes a structured, theoretically grounded map showing that the medical cannabis literature is heavily concentrated on clinician knowledge while neglecting social, environmental, and identity-based behavioural determinants. It does not establish which domains actually predict implementation behaviour, and its pre-2019 search window limits currency. For practice, the clearest implication is that knowledge-only educational programs are likely insufficient and that multi-domain approaches targeting peer norms, institutional support, and professional identity deserve both research attention and programmatic investment.

Why does my doctor seem uncomfortable discussing medical cannabis?

This review suggests that your doctor’s hesitation likely goes beyond simply not knowing enough about cannabis. Their professional training may not have included cannabis education, their colleagues may not discuss it openly, and the regulatory environment they work in may create uncertainty about what is legally and professionally appropriate. All of these factors contribute to the discomfort, and they take time to change even as the evidence base improves.

Will more education for doctors solve this problem?

Education is necessary but probably not sufficient on its own. This review found that the entire medical cannabis research literature has focused on knowledge gaps while largely ignoring the professional culture, peer norms, and workplace factors that also shape how clinicians behave. Effective change will likely require educational programs that address these broader influences alongside factual content.

Does this review tell us what works to change clinician behaviour around cannabis?

No. This is a mapping exercise that identifies where the research has focused and where important gaps exist. It does not test any intervention or measure what actually changes clinician behaviour. Its value lies in directing future research and educational design toward the right questions, but specific solutions still need to be developed and tested.

Have thoughts on this? Share it: