A large Australian survey of over 2,300 medicinal cannabis patients finds that those attending specialty cannabis clinics are younger, less medically complex, and consistently less satisfied with their care than those who receive prescriptions from their regular doctors. However, because patients chose their own treatment setting, the study cannot determine whether clinics themselves cause lower satisfaction or simply attract a different kind of patient.

Specialty cannabis clinics have become the dominant gateway to legal medicinal cannabis in Australia, with over 100,000 applications to the Therapeutic Goods Administration each year. Yet almost nothing is known about whether patients served by these clinics receive the same quality of care as those whose general practitioners prescribe cannabis as part of broader medical management. As similar models emerge in the United States, Canada, and Europe, understanding the tradeoffs between specialized access and integrated care is a pressing clinical and policy question that affects how regulators set standards and how patients make informed choices about where to seek treatment.

Australia’s medicinal cannabis landscape has shifted dramatically since legal access pathways were established, with specialty cannabis clinics now accounting for the vast majority of prescriptions. This study, drawn from the Cannabis As Medicine Survey 2022 (CAMS-22), is the largest direct comparison of patients attending these specialty clinics versus those receiving cannabis prescriptions from generalist health providers such as general practitioners or pain clinics. The researchers hypothesized that the two settings would serve distinct patient populations and generate different care experiences, and the survey was designed to capture demographics, health status (via the validated PROMIS-10 instrument), treatment characteristics, costs, and satisfaction across multiple care dimensions.

The findings confirmed that hypothesis emphatically. Medicinal cannabis clinic patients were approximately 3.5 years younger, more likely to be employed (OR 2.1), more likely to be treated for mental health conditions (OR 1.6), and more likely to use inhaled and THC-dominant products (OR for oral product use: 0.4). Generalist patients were older, had worse self-rated physical and mental health (PROMIS-10 scores roughly 2 to 3 T-score points lower), were more commonly treated for chronic pain, paid about $18 less per week, and reported higher satisfaction across every measured domain, including consultation duration, information quality, costs, and integration with other treatments. Perhaps most striking, only 3% of cannabis clinic patients had a prescriber who also managed their other health conditions, compared to 49% in generalist settings (OR 0.03). The authors appropriately note that the unadjusted, cross-sectional design and convenience sampling prevent causal conclusions, and they call for prospective studies with clinical outcome measures.

Care Fragmentation in Cannabis Medicine: What a 2,394-Person Survey Reveals

In less than a decade, Australia went from almost no patients with legal cannabis prescriptions to over a hundred thousand TGA applications per year, and the engine of that transformation has been a new kind of medical service: the cannabis-only clinic. But what do patients actually think of these clinics, and are they getting the same quality of care as those whose regular doctors prescribe their cannabis? This study, covering nearly 2,400 patients, attempts to answer that question with the broadest Australian dataset we have. What it genuinely contributes, before any criticism, is a clear and data-rich portrait of two strikingly different patient populations. Cannabis clinic patients are younger, healthier, more employed, and more likely to be using THC-dominant inhaled products for mental health conditions. Generalist patients are older, sicker, more often managing chronic pain, and far more likely to have a prescriber who also oversees their broader medical care. That last finding, an odds ratio of 0.03 for integrated care at cannabis clinics, is the most defensible and clinically important number in this paper. It is not ambiguous. It tells us that the cannabis clinic model, as currently practiced, is structurally isolated from the rest of a patient’s healthcare.

The satisfaction gap, however, requires much more careful handling. The paper reports that generalist patients were more satisfied across every dimension surveyed, and this headline finding is likely to be cited as evidence that cannabis clinics deliver inferior care. But the central methodological problem is that the analysis never adjusts for the profound baseline differences between groups. Every satisfaction comparison uses a single-covariate model: setting type alone, with no correction for age, health status, condition type, or prior cannabis experience. To put it plainly, this is the equivalent of comparing satisfaction ratings at a specialty running shoe store versus a department store without accounting for the fact that marathon runners tend to be harder to please than casual shoppers. The question is not just “which store scored higher?” but “are these the same customers?” Here, they clearly are not. Generalist patients may be more satisfied in part because they already have a trusted relationship with their doctor, because their expectations are different, or because their health conditions are managed more holistically. Without multivariable adjustment, we simply cannot separate the effect of the setting from the effect of who walks through the door.

What I would tell a patient is straightforward: if you are using a cannabis clinic, make sure your regular GP knows what you are taking, because the data suggest cannabis clinics rarely coordinate with your other care. What I would tell a colleague is that the integrated care finding deserves serious attention regardless of the satisfaction data’s limitations, because a 3% rate of co-managed care is a structural problem, not a statistical curiosity. And what I would tell a policymaker is that this paper justifies minimum standards for documentation and referral at cannabis clinics, but should not be cited as proof of inferior quality until prospective, outcome-anchored, multivariable studies confirm or refute the satisfaction patterns. In any comparison of healthcare settings using patient satisfaction data, the most important question to ask first is: are these the same patients? When they are not, as is unmistakably the case here, satisfaction differences tell you as much about who attends each setting as about how well each setting performs.

This study sits at the descriptive, hypothesis-generating stage of a research arc that has barely begun. We now have a detailed snapshot of who attends which type of cannabis prescribing service in Australia, and that is genuinely useful for planning further studies. But we have no prospective data, no clinical outcome measures, no objective consultation metrics, and no multivariable adjustment to separate patient-mix effects from setting effects. The research agenda the paper calls for is exactly right: cohort studies with linked patient registry data, clinical outcome tracking, and analyses that can control for the profound confounders documented here.

From a pharmacological and safety standpoint, the observation that cannabis clinic patients are more likely to use THC-dominant inhaled products and have higher rates of moderate-to-severe cannabis use disorder (14% versus 9%) warrants monitoring, though causality cannot be established. Clinicians should note that the near-total absence of integrated care at cannabis clinics raises specific concerns about drug interactions, polypharmacy oversight, and continuity of care for patients with comorbid conditions. The single most actionable recommendation from this evidence is to ensure that any patient receiving a cannabis prescription from a specialty clinic also has a generalist provider who is informed about, and involved in, their overall treatment plan.

This is a cross-sectional online survey using convenience sampling, which occupies a lower tier of the evidence hierarchy. It can describe associations between treatment settings and patient characteristics at a single point in time, but it cannot establish causal relationships. The single most important inference constraint is that patients self-selected into both the survey and their treatment setting, meaning unmeasured confounders almost certainly influence every reported association between clinic type and satisfaction.

This study extends earlier iterations of the Australian Cannabis As Medicine Survey (CAMS-18 and CAMS-20), which documented the growing dominance of specialty cannabis clinics in the prescribing landscape but did not directly compare patient profiles and satisfaction by setting. The finding that generalist providers offer more integrated care aligns with broader primary care literature showing that patients managed by a single coordinating provider report better care experiences. Dobson and colleagues have separately documented practitioner ambivalence toward cannabis clinic prescribers, suggesting that the structural isolation observed here may also reflect professional tensions. No prior Australian study has quantified the care fragmentation gap as starkly as this one does, making it the most specific data point available for regulators considering quality standards for cannabis-specific services.

Yes, and substantially. The most consequential analytic choice in this paper is the exclusive use of single-covariate Bayesian regression models, meaning treatment setting is the only predictor in each analysis. If the authors had run multivariable models adjusting for age, health status (PROMIS-10 scores), primary condition, and employment, the satisfaction differences between settings could have shrunk considerably or even reversed. Given that generalist patients were older, sicker, and treated for different conditions, the unadjusted satisfaction gap almost certainly captures a mixture of true setting effects and patient-mix effects. A sensitivity analysis using propensity-score matching or inverse probability weighting would have provided a much more credible estimate of the independent association between setting type and satisfaction.

The most likely and most consequential misreading of this paper is the conclusion that cannabis clinics provide worse care than generalist physicians. The study measures patient-reported satisfaction, not objective care quality, and the two are not interchangeable, particularly when the patient populations differ so dramatically. Satisfaction ratings are shaped by expectations, familiarity with the provider, health burden, and the specific conditions being treated. A younger, healthier patient seeking mental health treatment at a new clinic may hold different expectations than an older chronic pain patient with a longstanding GP relationship. Until multivariable analyses or prospective studies with clinical outcomes disentangle these factors, the satisfaction gap should be interpreted as a signal worth investigating, not as a settled verdict on care quality.

This study establishes that Australian cannabis clinic patients differ substantially from generalist patients in demographics, health status, and treatment patterns, and it documents a near-total absence of integrated care at cannabis clinics. It does not establish that cannabis clinics deliver inferior care, because its unadjusted, cross-sectional design cannot separate setting effects from patient-mix effects. For now, the most defensible clinical action is to ensure patients using cannabis clinics maintain active engagement with a generalist provider.

Does this study prove that cannabis clinics are worse than regular doctors?

No. The study shows that patients at cannabis clinics reported lower satisfaction, but the patients at each type of setting were very different in age, health, and the conditions they were being treated for. Without accounting for those differences, we cannot know whether the satisfaction gap reflects the quality of the clinic or the characteristics of the people who attend it.

Should I switch from a cannabis clinic to my regular doctor?

That depends on your circumstances. What this study most clearly shows is that almost no cannabis clinic patients had a prescriber who also managed their other health conditions. If you use a cannabis clinic, the most important step is to make sure your regular GP knows about your cannabis prescription and can coordinate your overall care.

Why were cannabis clinic patients younger and healthier?

The study cannot fully answer this, but likely explanations include that younger patients may find cannabis clinics through social media and online marketing, that older patients with established GP relationships may prefer to add cannabis to their existing care, and that cannabis clinics may particularly attract patients seeking treatment for mental health conditions rather than chronic pain.

Is this relevant to patients outside Australia?

The specific regulatory context is Australian, but the underlying tension between specialty access and integrated care applies to any country where cannabis-only clinics operate alongside generalist prescribers. The finding about care fragmentation is likely relevant wherever similar service models exist.

Have thoughts on this? Share it: