A 2018 survey of 171 medical cannabis recommenders found that most were self-taught, fewer than one-third received cannabis education in formal medical training, and nearly three-quarters of licensed providers had recommended cannabis as a substitute for prescription drugs. These practices are developing entirely outside the evidence-based guidelines that govern other areas of clinical care.

Millions of patients in the United States and abroad now access medical cannabis under the guidance of clinicians and unlicensed consultants who largely trained themselves. No formal clinical practice guidelines govern cannabis dosing, drug interactions, or contraindications, yet providers are routinely making high-stakes decisions, including recommending that patients replace prescription medications with cannabis. Understanding who these providers are and how they practice is foundational to assessing whether current patients are being adequately served or quietly exposed to uncharacterized risk.

Medical cannabis use has expanded rapidly across jurisdictions, yet the clinicians and consultants guiding patients through product selection, dosing, and administration have remained largely invisible to systematic study. This cross-sectional survey, conducted via the REDCap platform between June and December 2018, recruited 171 respondents from cannabis professional associations and dispensary networks across 22 U.S. states and 12 countries. Respondents included 144 licensed healthcare providers, among whom registered nurses comprised the largest group (46.5%), and 25 unlicensed individuals who served as cannabis “provider proxies.” The study aimed to characterize education sources, clinical indications, dosing strategies, and administration preferences, with a secondary comparison between licensed and unlicensed recommenders.

The survey found that while 93.1% of respondents reported some cannabis-specific education, fewer than one-third of licensed providers received endocannabinoid system training through formal medical or nursing education. Licensed providers were more likely to report formal cannabis training than unlicensed respondents (95.5% vs 76.9%; OR 6.3, 95% CI 1.2 to 32.3), but this difference did not survive Bonferroni correction for multiple comparisons. Nearly 72% of licensed respondents reported recommending cannabis as a substitute for an existing prescription medication. Vaporization was the preferred delivery route when THC was the therapeutic target, while oral and sublingual routes dominated for CBD. “Start low, go slow” titration emerged as the predominant dosing philosophy. The authors acknowledge that their convenience sample recruited through advocacy organizations limits generalizability and call for research freed from Schedule I barriers.

Flying Without Instruments: How Medical Cannabis Providers Are Guiding Patients Without Clinical Roadmaps

Imagine a medical specialty where the practitioners largely trained themselves, where no clinical practice guidelines exist, where nearly three-quarters of clinicians are recommending that patients swap their prescription medications for an unregulated botanical, and where the entire enterprise is invisible to the medical establishment overseeing every other domain of clinical care. That specialty exists. It is medical cannabis recommendation, and a 2018 survey by Corroon, Sexton, and Bradley offers our first detailed look inside. The paper claims to characterize the education, practices, and administration preferences of medical cannabis recommenders, and it does accomplish this within its own sample. What it actually tested, though, is far narrower: it surveyed a self-selected group of 171 individuals recruited through cannabis professional associations and advocacy networks, then described what those individuals say they do. This is a meaningful contribution. Before this study, there was essentially no empirical documentation of how cannabis providers make clinical decisions. The granular data on route-of-administration preferences by cannabinoid type, the emergence of “start low, go slow” as a near-universal titration norm, and the striking finding that 72% of licensed respondents recommend cannabis as a prescription substitute are all valuable observations that deserve attention from educators and regulators. That said, the central methodological problem is severe: recruiting respondents from cannabis advocacy organizations is like surveying members of the American Heart Association to characterize the “average” cardiologist. You will systematically overrepresent the most engaged, educated, and guideline-conscious practitioners and miss entirely the most casual, underqualified, or potentially harmful ones. The picture that emerges is almost certainly more optimistic than reality.

This matters because the paper’s implicit message, that a functional community standard of practice is forming, carries reassuring undertones that the sampling cannot support. The providers least likely to join professional associations, attend conferences, or sit through a 255-question survey are precisely the ones most likely to give patients inadequate or dangerous guidance. Furthermore, the “perceived effectiveness” scores reported throughout the paper reflect provider beliefs rather than patient outcomes. To put it plainly: asking a true believer to rate how well their treatment works is like asking a restaurant owner to review their own food. The feedback is real, but the benchmark is not independent. The study also cannot address perhaps the most urgent question it raises: when licensed clinicians recommend cannabis as a substitute for prescription drugs, including opioids, benzodiazepines, and antidepressants, what happens to those patients? Are drug interactions being monitored? Are tapering protocols in place? The data do not and cannot answer this. Similarly, the 15% of respondents who were unlicensed “provider proxies” represent individuals operating entirely outside healthcare accountability systems. The paper flags this concern but the survey was not designed to assess how these individuals’ recommendations differ in quality or safety from those of licensed clinicians.

What I would tell a patient is straightforward: if you are seeking cannabis guidance, know that most providers in this space are largely self-taught and that recommendations can vary enormously. The advice to replace a prescription medication with cannabis carries real but unquantified risks. To a colleague, I would say this study confirms our clinical intuition, that a cannabis recommender community has matured far enough to develop informal norms and professional infrastructure, but that this is not the same as validated practice. To a policymaker, the message is even more direct: this survey documents a structural gap that demands mandatory cannabis education in healthcare training, a viable pathway for rigorous research freed from Schedule I constraints, and oversight standards for unlicensed consultants. The paper’s most durable lesson is this: when a treatment enters widespread clinical use before the evidentiary infrastructure exists to evaluate it, providers do not stop practicing. They improvise. Documenting that improvisation, as this study does, is the essential first step. Mistaking the map for the territory, treating described practices as validated ones, is the critical error to avoid.

This study sits at the very earliest stage of the clinical research arc for medical cannabis practice characterization. It is hypothesis-generating, not hypothesis-testing. No randomized trials, prospective cohorts, or pharmacovigilance registries exist to contextualize or validate the practices described here. The finding that “start low, go slow” titration is broadly adopted suggests that informal consensus-building is underway, but whether this norm produces better patient outcomes than any alternative remains entirely unknown. The study’s 2018 data also precede substantial regulatory and market changes, limiting its timeliness for current practice.

From a pharmacological perspective, the most concerning finding is the prevalence of prescription substitution. Cannabis interacts with cytochrome P450 enzymes, particularly CYP3A4 and CYP2C19, and can alter the metabolism of commonly co-prescribed medications including warfarin, clobazam, and certain antidepressants. The survey provides no data on whether providers are monitoring for these interactions or counseling patients about tapering risks. Clinicians should recognize that any patient presenting with a cannabis recommendation to replace a prescription drug requires the same pharmacovigilance attention as any other therapeutic transition, including baseline labs, interaction screening, and structured follow-up.

This is a cross-sectional descriptive survey that collected self-reported data from a convenience sample at a single point in time. It sits near the base of the evidence hierarchy, generating descriptive statistics and limited bivariate comparisons but incapable of establishing causal relationships or assessing patient outcomes. The single most important constraint on inference is the non-random sampling from cannabis-advocacy organizations, which means these findings cannot be generalized to the broader population of individuals who recommend medical cannabis.

Prior surveys of medical students and family physicians in states like Colorado have documented substantial gaps in endocannabinoid system knowledge and cannabis-specific clinical training, consistent with the findings here that fewer than one-third of licensed providers received formal cannabis education in their professional programs. Separate work examining dispensary staff practices has similarly highlighted the role of non-clinician advisors in shaping patient behavior, a theme this study extends by including unlicensed “provider proxies” in its sample. What distinguishes this study is its granular attention to route-of-administration preferences stratified by cannabinoid type and its documentation of prescription substitution rates, neither of which had been systematically quantified before. The finding that vaporization predominates for THC while oral and sublingual routes are preferred for CBD aligns with pharmacokinetic logic but has not been tested in comparative effectiveness research.

The most consequential analytic choice was the application of Bonferroni correction to the primary comparison between licensed and unlicensed providers. Without it, the odds ratio of 6.3 for formal cannabis education favoring licensed providers was statistically significant at p = 0.03. With correction, it was not. This is appropriate given the multiple comparisons conducted, but it also means the paper’s headline distinction between licensed and unlicensed provider education is not statistically supported. Had the authors pre-specified a smaller number of primary comparisons or used a less conservative correction method such as false discovery rate control, the finding might have retained significance. Additionally, a population-weighted sampling approach using state cannabis authorization registries rather than advocacy networks would have produced a substantially different, and likely less favorable, portrait of provider education and practice quality.

The most likely overinterpretation is reading this study as evidence that medical cannabis providers are adequately trained and that patients are being safely guided. The data show only what a self-selected group of advocacy-aligned respondents report about their own practices. Providers who are less educated, less engaged, or less scrupulous are systematically absent from this sample. Similarly, the detailed perceived-effectiveness ratings reported in the paper reflect provider beliefs, not measured patient outcomes, and should not be cited as evidence that cannabis works for any specific condition. The finding that 72% of licensed providers recommend cannabis as a prescription substitute is descriptive, not evaluative. It tells us the practice exists at scale but says nothing about whether it is safe, effective, or harmful.

This study contributes the first granular empirical description of how medical cannabis recommenders practice, including their education sources, dosing philosophies, route preferences, and prescription substitution habits. It does not establish whether these practices are safe, effective, or representative of all cannabis providers. For clinical practice today, it serves as a clear signal that the evidence infrastructure surrounding medical cannabis guidance has not kept pace with its real-world use, and that patients, providers, and regulators are all operating in the same information vacuum.

Does this study prove that medical cannabis providers are well-trained?

No. While most respondents reported some cannabis-specific education, fewer than one-third received it through formal medical training. The survey also recruited from advocacy organizations, meaning participants are likely more educated and engaged than the average cannabis recommender. The study cannot speak to the training quality of providers outside this self-selected sample.

Is it safe to replace my prescription medication with cannabis based on a provider’s recommendation?

This study does not address safety. It found that nearly three-quarters of licensed providers have recommended such substitutions, but it did not measure patient outcomes, adverse events, or drug interactions. Any transition from a prescription medication to cannabis should involve a licensed clinician who can monitor for interactions, manage tapering, and conduct appropriate follow-up.

What does “start low, go slow” mean for cannabis dosing?

“Start low, go slow” is an informal dosing approach in which patients begin with a very small amount of cannabis and gradually increase the dose over days or weeks until they achieve the desired therapeutic effect or encounter side effects. This approach was the most commonly reported strategy in the survey. While it reflects a reasonable harm-reduction philosophy, it has not been validated in clinical trials and does not constitute an evidence-based dosing protocol.

Should I be concerned if my cannabis advisor is not a licensed healthcare provider?

This survey found that about 15% of respondents were unlicensed individuals acting as cannabis consultants. These individuals operate outside the regulatory oversight, malpractice accountability, and continuing education requirements that apply to licensed clinicians. If you are using cannabis for a medical condition, it is advisable to work with a licensed healthcare provider who can integrate cannabis guidance with your overall medical care.

Have thoughts on this? Share it: