Clinician Beliefs and Practices Related to Cannabis

Introduction

Cannabis with ≥0.3% delta-9 tetrahydrocannabinol [THC] concentration (also called marijuana ^* ) remains a schedule I substance in the United States under the 1970 U.S. Controlled Substances Act. ¹ By definition, this means that it has no currently accepted medicinal use in treatment in the United States, lacks accepted safety data for use under medical supervision, and has high potential for abuse. However, as of November 2020, cannabis has been legalized for medical use in 36 states and the District of Columbia (DC); 15 states and DC have also legalized nonmedical adult cannabis use. ² All but three of the remaining states have legalized access to hemp, low-THC, or cannabidiol (CBD)-only cannabis products, ² which are typically used medically.^3,4 While these state-based policies are relatively recent, cannabis has documented use (medicinally) dating back >5000 years ⁵ and was first described in the U.S. Pharmacopeia for medicinal use in 1850 (and subsequently used as a patented U.S. medicine during the 19th and early 20th centuries). ⁶ Currently, the U.S. Food and Drug Administration has approved three synthetic cannabis-related drugs (containing laboratory-derived, isolated THC compounds) and one plant-derived cannabinoid drug for indicated medical treatments. ⁷

Despite recent state policy changes, the science is still emerging around the health effects of cannabis. The 2017 National Academies of Sciences, Engineering, and Medicine (NASEM) reviewed this scientific evidence and concluded that there is substantial evidence that cannabis and cannabinoids are effective for treatment of chronic pain in adults, as antiemetics in the treatment of chemotherapy-induced nausea and vomiting, and for improving patient-reported multiple sclerosis spasticity. ⁸ Of note, some of the evidence reviewed in the NASEM report is based on studies of isolated cannabinoid medications—both synthetic and plant derived. ⁸ The report also found moderate evidence for improved short-term sleep outcomes in individuals with sleep disturbance. ⁸ However, these therapeutic effects are accompanied by potential health risks, including chronic bronchitis and increased respiratory symptoms (primarily due to smoked or inhaled cannabis); short-term impairment in learning, memory, and attention; impaired driving and risk for motor vehicle crashes; lower birth weight among babies born to mothers who used cannabis; and increased risk for the development of schizophrenia and other psychoses, cannabis use disorder, and other substance use disorders.^8,9 A number of these risks increase with earlier initiation and more frequent/heavy use patterns. ⁸

Clinicians play an important role in translating this science into practice and weighing these benefits and risks. In states with legal medical cannabis use, clinicians also play an important role in providing patients access by recommending cannabis when clinically indicated. ¹⁰ However, little is known about clinician knowledge, attitudes, and behaviors related to medical cannabis. A recent study of a convenience sample of 62 primary care providers in a large Minnesota-based health care system ¹¹ found that more than half of the clinicians believed that cannabis had medical uses, over one-third believed clinicians should be offering cannabis to patients for certain medical conditions, and nearly three-quarters of clinicians were interested in learning more about medical cannabis. ¹¹ Another study of a nationally representative sample of 400 oncologists found that only 30% felt informed to make recommendations about medical cannabis, yet 80% discussed it with patients and 46% recommended it clinically. ¹² A study of 494 clinicals in Washington State found that 27% had recommended cannabis despite having a legal medical and nonmedical cannabis market in the state. ¹³ A majority of clinicians reported low levels of knowledge and comfort in recommending cannabis. ¹³ A 2013 study of 520 family physicians in Colorado found that 46% did not support physicians recommending medical cannabis and that a minority thought cannabis conferred significant benefits for physical health (27%) and mental health (15%). ¹⁴ Other studies have focused on health professionals treating specific patients (e.g., end-of-life patients and pediatric patients)^15,16 or on clinicians outside of the United States.^17,18 No multistate multispecialty studies on this topic have been published in academic journals.

Understanding clinician beliefs and behaviors related to medical cannabis can identify gaps and inaccuracies in clinician knowledge about medical cannabis and can inform future clinician education and training. Accordingly, this study sought to explore beliefs and behaviors related to medical cannabis in a sample of family practice doctors, internists, nurse practitioners, and oncologists.

Methods

Results

Overall, a higher proportion of the sample were male (57.9%), 41–60 years old (57.2%), and in group outpatient practice (70.6%). The mean number of years of practice was 17.3 with a range of 3 to 53 years (Table 1). In terms of specialty, the majority of clinicians were general practitioners: 32.1% of the sample were family practice doctors, 34.4% were internists, 16.7% were nurse practitioners, and 16.8% were oncologists. Overall, the majority (75.4%) of clinicians reported asking patients about cannabis (marijuana) use and just over half (53.5%) reported assessing patients for cannabis use disorder. By clinical specialty, the prevalence of asking about cannabis use (82.1%) was highest among nurse practitioners and lowest among oncologists (60.1%); and the prevalence of assessing for cannabis use disorder (61.3%) was highest among family practice clinicians and lowest among oncologists (30.8%) (Table 1).

Overall, 68.9% of clinicians surveyed believed that cannabis (marijuana) has medicinal uses, with the highest prevalence of belief in medical uses (80.2%) among nurse practitioners. Among those who believed that cannabis had medicinal uses, a majority of clinicians surveyed believed that cannabis had medicinal use for pain (73.1%), cancer (71.7%), nausea (60.5%), and appetite activation (60.1%). These were the most frequently endorsed conditions across all clinician types. Conditions with the lowest prevalence of clinicians endorsing medicinal use were opioid addiction (29.7%), ADHD (13.6%), and Alzheimer's disease (12.9%) (Table 1). These patterns were generally similar across clinician specialty groups. Overall, 26.6% of clinicians indicated that they had recommended cannabis to a patient, with 27.3% of family practice doctors, 24.9% of internists, 22.6% of nurse practitioners, and 32.4% of oncologists reporting that they had recommended cannabis.

Overall, 15.3% of clinicians reported that they were in a state with approved nonmedical cannabis legalization, 24.2% reported that they were in a state with legalization for medical cannabis only, 27.6% reported that they were in a state with medical legalization of low-THC products only, and 32.9% reported that they were in a state with no cannabis legalization policies in place (Table 1). When we coded cannabis legalization policies by state and assessed the accuracy of clinicians' self-reported state cannabis legalization policy, 60.0% of clinicians had incorrectly reported the cannabis legalization policy in their state (Table 2). Notably, analyses showed that 22.1% of clinicians indicated that they were in a state with no policies legalizing cannabis, but were actually in a state with policies legalizing low-THC/CBD products; 21.0% of clinicians indicated that they were in a state with only low-THC/CBD products, but were actually in a state with medical legalization; and 9.5% of clinicians indicated that they were in a state with no policies legalizing cannabis or cannabinoids, but were actually in states with medical legalization. In multivariable analyses, the only significant predictor of accurately reporting a state's cannabis policy was ever having recommended cannabis to patients (adjusted odds ratio 2.2, confidence interval [95% CI] 1.7–2.8) (data not shown).

In multivariable analyses, significant correlates of asking patients about cannabis use included being female (vs. male), being 40 years of age or younger (vs. older than 60 years), and being a family practice doctor, internist, or nurse practitioner (compared with being an oncologist) (Table 3). Those working in individual outpatient settings had lower odds of asking about cannabis than those in inpatient practice. Correlates of believing cannabis has medicinal use (vs. no medicinal use or not knowing) included being a nurse practitioner (vs. an oncologist). In terms of correlates of ever recommending cannabis to patients, internists had lower odds of ever recommending cannabis compared with oncologists; no other correlates were significant (Table 3). Those who believed cannabis had medicinal use had 5.9 times the odds (95% CI 3.9–8.9) of recommending cannabis to patients compared with those who were unsure about the medical uses of cannabis (Table 4).

Discussion

This is among the first studies to assess clinician beliefs and practices related to medical cannabis in a U.S. multistate sample. Important findings from this study include the following: (1) over two-thirds of clinicians surveyed believed that cannabis (marijuana) had medicinal uses, although just over a quarter had ever recommended cannabis to a patient; (2) beliefs about conditions for which cannabis can be used medically did not necessarily align with the current scientific evidence; (3) more than half of the clinicians surveyed incorrectly reported the legal status of cannabis in their state; and (4) while over three-quarters of clinicians ask patients about cannabis use, only about half assess patients for cannabis dependence or use disorder.

Across all surveyed clinician types, believing cannabis had medical use was strongly correlated with ever recommending cannabis to patients. However, a much higher proportion of clinicians believed that cannabis had medical uses than clinicians who had ever recommended cannabis to patients. While asking follow-up questions to determine reasons for this discrepancy was beyond the scope of the current study, there are a number of possible reasons for this. First, clinicians may lack adequate knowledge about cannabis to feel comfortable making a recommendation.^10,11 Second, obtaining information about dosing and product types (as clinicians do for other medicines) may be difficult for many doctors given that plant-based cannabis is not an FDA-regulated medicine, ²⁰ and many medical cannabis programs in U.S. states are set up to have a medical cannabis dispensary facilitate those aspects of use.^10,21 While some states require clinicians or pharmacists to be at the dispensary,^22,23 communication between those individuals and recommending clinicians could be lacking. In the majority of states, dispensary employees (often called budtenders) act as a proxy for clinicians despite having little or no medical training. ¹⁰

While the same general discrepancy existed between beliefs and practices related to medical cannabis, some variation existed across specialties. For example, of the clinical specialties surveyed, the highest prevalence of belief in medicinal cannabis use was found among nurse practitioners, but with the lowest prevalence of ever recommending cannabis. This may be due to differences in state medical cannabis policies in terms of who can recommend or endorse medical cannabis use—with some states allowing only physicians to recommend use^24–26 and others allowing a broader range of clinicians.^27,28

Another barrier to recommending cannabis despite believing in its medicinal use may be that clinicians are unaware of the current scientifically backed uses for cannabis. ¹¹ Results from this study suggest that the highest prevalence conditions where clinicians indicated they believed cannabis could be medically used were scientifically based—pain, nausea, appetite activation, antiseizure, and spasticity^8,9—and this was generally true across clinician specialties. However, a number of conditions that were endorsed by clinicians lack scientific evidence for current use. For example, nearly half of the clinicians believed that cannabis could be used medically for glaucoma, an area where little research exists and findings have been inconclusive. ⁸ Similarly, over one-third of clinicians believed that cannabis could be used medically for depression or anxiety when some evidence actually indicates that cannabis use (especially heavy marijuana use) could be harmful for depression, particularly if suicide ideation exists.^8,29,30 Furthermore, almost one-third of clinicians believed that cannabis could be used medically for opioid use disorder, which is not supported by existing scientific evidence.^31–33 These findings may be a by-product of state authorizations for medical cannabis use in certain conditions. ³³ For a variety of reasons, some states have authorized medical cannabis use for conditions that are not yet well supported by the science.^8,34 Furthermore, there are a number of instances where the scientific research on medicinal uses of cannabis is incomplete or unclear, which could present challenges to clinical practice. Findings from this study suggest that translating the science and providing education to clinicians about potential benefits and risks of cannabis are warranted.

Clinician education about state-based policies for cannabis use may also be warranted. In this study, 6 in 10 clinicians incorrectly reported the cannabis legalization policy in their state. Over a quarter of clinicians reported that they had nonmedical (recreational) legalization policies in their state when they did not. Almost a quarter reported that their state had access to medical cannabis when the state actually had a policy in place for access to low-THC or CBD-only products. Clinicians' knowledge of state cannabis policy is important in terms of patient access (or lack thereof) and state oversight and regulation of available products given that cannabis with ≥0.3% delta-9 THC concentration remains illegal at the federal level. In particular, although it may seem nuanced, clinicians' understanding of the differences in policies that legalize low-THC or CBD products versus those that legalize medical cannabis products is important as potential legal ramifications can accompany these differences.

Last, opportunities remain for clinical education to improve practices for asking about and screening for cannabis use and use disorder in patients. While a majority of clinicians reported asking patients about cannabis use, just over half reported assessing for cannabis use disorder. Substantial evidence exists that using cannabis can be associated with cannabis use disorder ⁸ particularly in cases of heavy and frequent use. ⁸ Even in cases of medicinal use, development of a cannabis use disorder is possible. The increasing availability of high-potency concentrates in the legal and illegal marketplaces may further increase the potential for cannabis use disorders. ³⁵

This study is subject to at least four limitations. First, data are derived from a web-based panel that draws a convenience sample; probability-based sampling approaches were not utilized. While being a multistate sample, data are not nationally representative and may not be generalizable. Second, not all clinician specialties are represented in these data, and data may differ across other specialties and subspecialties. Third, data are cross-sectional and from 2018. These data cannot be used to measure changing beliefs among clinicians and practices over time. It is possible that clinicians' beliefs and behaviors may have evolved. Fourth, the survey instrument was limited in terms of questions that could be included. Questions did not assess reasons for specific beliefs or practices. Additional qualitative research may be warranted to better understand where clinicians derive information about cannabis, why they believe in its effectiveness for certain conditions, and how they make decisions about whether to recommend cannabis.

Findings from this study may have implications for clinician education and training across U.S. states. These data suggest that some clinicians believe that cannabis has medicinal use, but may not fully understand the science behind the health effects of cannabis, nor be willing to recommend cannabis to patients. In addition, findings suggest that clinicians may not have an accurate understanding of their state-based policies for cannabis legalization and use. Clinicians remain a critical group to educate about the health effects of cannabis and cannabis legalization policies because of their role in translating science into practice, weighing benefits and risks with patients, and recommending cannabis for medicinal use in states that have legalized it. Education about the health effects of cannabis and about clinician roles in the current cannabis policy landscape could occur as part of medical school, residency, or fellowship; as part of continuing education requirements; or as part of license renewal. Ongoing research to better understand the reasons for clinicians' beliefs and behaviors related to cannabis could help further inform these educational and training efforts.