Physicians’ experiences, attitudes, and beliefs towards medical cannabis: a systematic literature review

Background An increasing number of countries legalise the use of medical cannabis or allow it for a narrow range of medical conditions. Physicians, and often the patients’ general practitioner, play a major role in implementing this policy. Many of them, however, perceive a lack of evidence-based knowledge and are not confident with providing patients with medical cannabis. The objectives of this review are to synthesise findings about hospital physicians’ and GPs’ experiences, attitudes, and beliefs towards the use of medical cannabis with the purpose of identifying barriers and facilitators towards providing it to their patients. Methods Peer-reviewed articles addressing hospital physicians’ and GPs’ experiences, attitudes, and beliefs towards the use of medical cannabis were searched systematically in PubMed, Scopus, EMBASE, and the Cochrane Library. Results Twenty-one articles were included from five different countries in which the medical cannabis laws varied. The studied physicians experienced frequent inquiries about medical cannabis from their patients (49–95%), and between 10 and 95% of the physicians were willing to prescribe and/or provide it to the patients, depending on setting, specialty and experience among the physicians. This review found that physicians experienced in prescribing medical cannabis were more convinced of its benefits and less worried about adverse effects than non-experienced physicians. However, physicians specialized in addiction treatment and certain relevant indication areas seemed more sceptical compared to physicians in general. Nevertheless, physicians generally experienced a lack of knowledge of clinical effects including both beneficial and adverse effects. Conclusion This review indicates that GPs and hospital physicians from various specialties frequently experience patient demands for medical cannabis and to some degree show openness to using it, although there was a wide gap between studies in terms of willingness to provide. Hospital physicians and GPs’ experienced in prescribing are more convinced of effects and less worried of adverse effects. However, most physicians experience a lack of knowledge of beneficial effects, adverse effects and of how to advise patients, which may comprise barriers towards prescribing. More research, including larger studies with cohort designs and qualitative studies, is needed to further examine facilitators and barriers to physicians’ prescribing practices.

when prescribed or provided by healthcare professionals. In most of these countries, the cannabis is dispensed through pharmacies [1]. Other countries, including Denmark, Norway, Sweden, Poland, and the United Kingdom allow treatment with medical cannabis for a narrow range of medical conditions in patients where all other options of conventional treatment have been tried without reaching treatment targets [1,2]. Most commonly, a specialist with a specific license prescribes the cannabis products, and also pharmacies need a license to supply them [1].
Medical cannabis has been debated worldwide among physicians and decision-makers, and the use of it remains controversial [3]. Moreover, the quality of the evidence of potential benefit as well as adverse effects is low [4,5]. Furthermore, cannabis contains tetrahydrocannabinol (THC) which is the euphoric component in cannabis for recreational use [6], and for this reason it gives rise to concerns about abuse and addiction [5,7]. However, medical cannabis and cannabis for recreational use are different from each other, as medical cannabis is subject to stricter requirements than recreational cannabis in terms of therapeutic safety, cultivation and manufacturing [6]. Another main component used in some preparations of medical cannabis is cannabidiol (CBD) which is non-euphoric. Depending on the needs to address, THC and CBD is given to patients in controlled, carefully metered doses [8].
Generally, physicians play a major role in implementing regulatory policies on the use of medical cannabis, and specifically general practitioners (GPs) who are often the patients' first contact in healthcare systems and an ongoing coordinator of their treatment [9]. A recent systematic review study reviewed the existing literature concerning all types of health care professionals' personal beliefs, knowledge, and concerns regarding delivery and supply of medical cannabis to patients [10]. However, it is just as essential to focus on physicians' experiences with patients' demand for medical cannabis and whether they decide to provide it to them [11].
Hence, in order to fill this knowledge gap, the objectives of this review were to investigate hospital physicians' and GPs' experiences with patients' demand for medical cannabis and prescription practice, as well as their attitudes, and beliefs towards the use of medical cannabis with the purpose of identifying barriers and facilitators towards providing it to their patients.

Methods
In this review we followed the guidelines given by the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) Statement [12]. Peer-reviewed articles addressing hospital physicians' and GPs' experiences, attitudes, and beliefs towards medical cannabis were searched in the databases PubMed, Scopus, EMBASE, and the Cochrane Library. We searched databases through February 2019. The search strategy included search terms listed in Table 1.
The search strategy was adjusted for each specific database and its search mechanisms. Criteria for inclusion and exclusion are shown in Table 2. Articles were included if they, addressed physicians' experiences, attitudes, or beliefs towards use of medical cannabis, were written in English, Danish, Norwegian or Swedish and published after year 2000. Articles were excluded if they only examined the use of medical cannabis in children. Besides, no exclusion criteria on the type of medical cannabis products were applied, and therefore, this review defines medical cannabis as both magistral preparations and herbal products. Magistral preparations are medical drugs with active ingredients of the cannabis plant, compounded in pharmacies and prescribed by the doctor Table 1 Search strategy (PubMed)

Variable Search terms
Medical cannabis ("medical cannabis" OR "medical marihuana" OR "medical cannabidiol" OR "medical cannabinoids" OR "medicinal cannabis" OR "medicinal marihuana" OR "medicinal cannabidiol" OR "medicinal cannabinoids" OR "medical use of cannabis" OR "medical use of marijuana" OR "medical use of marihuana" OR "medical use of cannabidiol" OR "medical use of cannabinoids" OR "medicinal use of cannabis" OR "medicinal use of marijuana" OR "medicinal use of marihuana" OR "medicinal use of cannabidiol" OR "medicinal use of cannabinoids" OR "clinical use of cannabis" OR "clinical use of marijuana" OR "clinical use of marihuana" OR "clinical use of cannabidiol" OR "clinical use of cannabinoids" OR "therapeutic use of cannabis" OR "therapeutic use of marijuana" OR "therapeutic use of marihuana" OR "therapeutic use of cannabidiol" OR "therapeutic use of cannabinoids" OR "Prescription of cannabis" OR "prescription of marijuana" OR "prescription of marihuana" OR "prescription of cannabidiol" OR "prescription of cannabinoids" OR "ordination of cannabis" OR "ordination of marijuana" OR "ordination of marihuana" OR "ordination of cannabidiol" OR "ordination of cannabinoids") AND to specific patients who do not benefit sufficiently from authorized medicines [6]. Title and abstract were independently screened by one author, based on the stated exclusion criteria. The remaining articles were screened in full text by two authors. Disagreements were resolved by consensus or consulting with a senior researcher. Details of the literature search are shown in a flow diagram in Fig. 1.

Characteristics of identified literature
This review includes 21 studies in total; 19 quantitative survey studies using questionnaires and two qualitative studies using open-ended questionnaires and interviews. All the identified studies aimed to explore experiences and/or attitudes among physicians towards the use of medical cannabis, and about half of the studies also investigated their perceived knowledge or educational needs regarding medical cannabis. Some studies included multiple types of physicians while other studies only included one or few specialties. The studies were conducted in different countries, however most studies were conducted in the United States (see Table 3).
In the studies included, patient groups differed, and the laws on the use of medical cannabis varied. The literature details are shown in Table 4. Due to the heterogeneity in study populations as well as between patient groups and legislation, it was not possible to conduct a formal metaanalysis. Instead, we conducted a literature review, and the study results are presented as text.
As our objective was to synthesize all existing literature about hospital physicians' and GPs' experiences, attitudes, and beliefs towards the use of medical cannabis, we did not perform a critical appraisal of the individual studies included. However, small study populations and low response rates are considered to be compromising for the validity of the study findings.

Procedures and definitions of medical cannabis prescriptions in included studies
In the United States (US), Canada, Australia and Israel, medical cannabis was legal, but the physicians' role in facilitating access varied. In none of the studies, physicians could provide medical cannabis to patients directly. In Ireland medical cannabis was illegal. In the US in general, medical cannabis was illegal according to the federal law, yet an increasing number of states have state-wise legalised medical use of cannabis [1]. All US studies included were conducted in such states. Physicians in states having authorised the medical use of cannabis could certify or recommend that their patients had a qualifying medical condition allowing the use of cannabis for medical purposes, but could not actually issue a prescription. The only lawful ways to dispense it, is as part of a federally approved research program or through state laws, which may permit caregivers and/or other healthcare workers to manufacture and distribute cannabis preparations to patients [34]. At the time when the studies were conducted, Canadian physicians could sign a document attesting that all conventional treatments had been tried and provide information on daily dose and duration of validity. Health Canada officials should subsequently give their approval [32]. In Israel, the use of medical cannabis was legal in terms of a licensing procedure, which meant that physicians could sign a medical recommendation, which was then processed and acknowledged by the Ministry of Health [1].
The terms used for the physicians' procedure of facilitating access to medical cannabis are not consistent in the included articles; hence, regardless of the terms used Addressed physicians' experiences, attitudes, or beliefs towards use of medical cannabis.
-We included studies that addresses both physicians and other health care professionals and/or patients, if we could extract results for physicians or if 50% or more of the study population were physicians.
-We included studies if they addressed medical cannabis by involvement of physicians in the identification of the perceived need and use of medical cannabis on the patients.
Were written in English, Danish, Norwegian or Swedish -Because we were interested in international peer-reviewed scientific articles, we included articles in the Scandinavian languages as well, because all authors were able to understand it.

Were published after year 2000
-The year was chosen as a pragmatic cut off, by which articles would be discarded, due to contextual differences in the use of medical cannabis among health professionals and in the public. Research activity within this topic is new and current.

Articles were excluded from this review, if they;
Only examined the use of medical cannabis in children (< 18 years) -Children are physiologically different from adults, and the use of medical cannabis in their treatment may have severe consequences [13]. Hence, the beliefs and attitudes of physicians, regarding use of medical cannabis, may differ depending on the age of the population under investigation.
in the individual studies, 'the term 'provide' was chosen to be consistently used throughout this review, because it both relates to the issue of prescriptions as well as other ways of supplying patients with medical cannabis.

Physicians' experiences with patients receiving medical cannabis Experiences with patient inquiries and prescriptions
Physicians experience inquiries about medical cannabis from a variety of patients, and some physicians provide it to their patients. The proportion of physicians having experienced inquiries about medical cannabis from patients varies between 49 and 95% in the identified studies [18-20, 24, 27]. The percentage of physicians reporting to have provided cannabis varies from 10 to 95% [15, 16, 19, 20, 22, 26-28, 31, 33, 35]. Especially three studies conducted in Israel report high proportions of physicians experienced in prescribing medical cannabis, namely 48, 60 and 95% respectively [31,33,35].
Seventy eight percent of physicians, feel uncomfortable with indicating dosage, frequency, and method of administration of cannabis prescriptions [35]. Studies show significant associations between physicians' experiences with prescribing medical cannabis and their self-reported knowledge of it and confidence in prescribing it. Significantly higher proportions of physicians experienced in prescribing medical cannabis feel comfortable with providing it [26,28] and report themselves as having greater knowledge of medical cannabis compared to physicians who have never provided it [33]. Additionally, another Israeli study finds that 60% of physicians report that they would not be willing to provide medical cannabis, without the licensing procedure at the Ministry of Health. This means that the ministry needs to approve a medical recommendation signed by the physicians, before the medical cannabis can be provided to the patient [33].

Experiences with effects, adverse effects, and misuse
In general, physicians experience a lack of knowledge about medical cannabis (64-90%) including beneficial as well as adverse effects [16,20,24,25,27,28,31,35], and they do not feel confident using it in treatment of patients [24]. Despite this, many physicians (46-95%) still choose to provide it [16,24,31]. A qualitative interview study including physicians' experiences with effects and adverse effects from medical cannabis, reported that some physicians, including family physicians and oncologists describe positive impressions of how medical cannabis helps their patients, and they view it as useful. They claim to see more efficacy of medical cannabis in real life than proven in literature, and hands-on experiences are mentioned as having a crucial impact on their views and on their decisions of providing it [9]. Similarly, in two recently published questionnaire-based studies, high proportions of physicians experience positive effects (63-67%) and mild or no side effects (56%) when using medical cannabis in treatment of certain ailments [16,31]. The respondents in both of these studies are supportive of medical cannabis, and the authors concurrently conclude that the positive attitude may stem from the fact that the physicians' patients experience beneficial effects of treatment with the medical cannabis.

Attitudes towards prescription
There are various attitudes towards the prescription of medical cannabis among physicians. Some physicians argue that cannabis is a social and criminal matter which should not fall under the professional domain of medicine [9]. Other physicians accept cannabis as medicine for patients with a specific diagnosis [9]. .Legislation on the use of medical cannabis is widely discussed in the literature, and physicians are typically asked to report their opinion [21,23,29]. Studies generally show that significantly lower proportions of physicians with specialty, or other educational skills in addiction medicine, support the legal use of medical cannabis, compared to physicians with other specialties, including general practitioners (36% vs. 60%) [23,29]. Mathern et al. investigated attitudes towards the use of medical cannabis in epilepsy patients and showed that a minority of epileptologists and general neurologists supported the use of medical cannabis for this group of patients [21].

Beliefs in effects, adverse effects, and misuse
Different beliefs in effects and adverse effects following clinical use of medical cannabis are reported among physicians. Conventional medicine is often seen as the ideal, and physicians argue that medical cannabis fails to comply with the standards of biomedicine [9,30]. They point out the lack of scientific evidence of safety and efficacy [9,30]. However, some physicians agree on the limited evidence, but argue that other aspects of health care are likewise unsupported by evidence [9]. When physicians consider treatment of pain and palliative care where all conventional methods have been tried, they give much less weight to the lack of evidence and potential harms and are more inclined to support the use of medical cannabis [9,17,28,30,33,35]. Finally, some physicians are generally concerned that patients who request medical cannabis may actually want it for recreational use [27,30].
An American study reports that almost 90% of its study population (palliative and hospice care providers) believe that cannabis can be useful in the treatment of pain, nausea, and appetite loss, and more than half of them believe that adverse effects are the same or less problematic than conventional treatments when considering pain, appetite loss, nausea, sleep, and end-of-life care [17].
Furthermore, hospital physicians' and GPs' beliefs in effectiveness seem to depend on their experiences and educational skills. Research shows that GPs experienced              in providing medical cannabis are more convinced about benefits and less worried about harmful adverse effects compared to physicians who have never provided it (see Table 4) [22]. Accordingly, a recent American study find a high proportion of physicians willing to provide medical cannabis if it was legal (71% as an oral form, 86% as a topical treatment) [18]. An Irish study shows significant associations between GPs' knowledge of substance misuse and their beliefs about efficiency, as a significantly smaller proportion of physicians with special training in addiction treatment believe that medical cannabis plays a role in pain treatment and palliative care, compared to physicians without this specialist training (see Table 4) [29].

Statement of principal findings
This review shows that GPs and hospital physicians in various specialties often experience inquiries about medical cannabis from their patients and that they are willing to provide it to some degree. Although it should be noted that the number of prescribing physicians varies considerably, depending on setting, specialty, and experience among the physicians. However, physicians generally experience a lack of knowledge of clinical effects, both beneficial and adverse effects. Regarding physicians' attitudes to and beliefs in medical cannabis, this review shows that physicians experienced in prescribing medical cannabis are more convinced of its benefits and less worried about adverse effects than physicians without experience. However, physicians specialized in addiction treatment and certain relevant indication areas seem to be more sceptical about using it for treatment of patients, compared to physicians in general.

Strengths and weaknesses of the study
The strength of this review lies in the systematic approach to identifying peer-reviewed studies. However, most of the included studies were small and limited by relatively low response rates, which compromises the validity of the study findings (Tables 3 and 4). It is possible that responders differ from non-responders, as responders may be more interested in the topic and have more knowledge of it than non-responders. Besides, nonresponders may have chosen not to participate, because they do not provide medical cannabis. This limits the generalisability of results of the single studies as well as this review. Consequently, the heterogeneous study populations and patient groups limit the comparability of results. Experiences with and attitudes towards medical cannabis may vary as to specialty and the type of patients whom physicians see and treat in their daily work life, which makes study results less comparable. On the other hand, the heterogeneous sample also allows for the presentation of apparent differences between specialties and physicians. Yet, it could also be considered as a strength that the focus is on hospital physicians and GPs alone, unlike a recent review encompassing many types of healthcare professionals regardless of substantial differences in their right to make decisions about treatment of patients [10]. However, the laws on prescription of medical cannabis vary among the studied populations and between countries. This may affect the results between the individual studies and make them less comparable. Additionally, all the quantitative studies included used cross-sectional designs, which limits our possibility of concluding on causality, including facilitators and barriers, to prescribing medical cannabis.
The restrictions to the language of eligible publications may potentially have limited our study results, as articles published in other languages may present experiences and attitudes from physicians that cannot be generalized to the countries that were represented in this review. However, due to limited resources we were not able to translate and include these articles.

Meaning of the study
This review shows that hospital physicians and GPs experienced in prescribing medical cannabis are more convinced about its benefits and less worried about adverse effects. This indicates that physicians have provided cannabis because they are confident about its effects. It is, however, possible that physicians to some degree report to be convinced of effects to justify their prescription. Furthermore, positive hands-on experiences with providing medical cannabis were described as having a crucial impact on their views, and the physicians' experiences may be a facilitator for providing medical cannabis.
Generally, physicians with specialty in epileptology and neurology did not support treatment of epilepsy patients with medical cannabis [21], and those with specialty in addiction treatment or similar indication areas seemed to be more sceptical compared to other medical professionals [29]. This indicates that hospital physicians' and GPs' attitudes to and beliefs in medical cannabis are associated with their specialty, which might be explained by their different perspectives and experiences with patients' needs, as well as their responsibilities for specific patient groups, e.g. patients who have substance abuse problems versus alleviation of pain in palliative patients. Physicians specialized in addiction treatment may mainly experience the adverse health effects from recreational use of cannabis, which possibly gives rise to their scepticism. However, these results are based on very few and small studies, and to gain deeper and more substantiated knowledge of possible associations between speciality training and prescription of medical cannabis more research is needed.
Regarding attitudes and beliefs, the results of this study are in line with another recent review [10]. However, this study adds new knowledge with its additional focus on actual experiences with providing it to patients. Hence, high proportions of hospital physicians and GP's experienced in prescribing medical cannabis were found in the Israeli studies, which might be explained by the fact that Israel has had a less restrictive legal attitude towards cannabis use compared with the USA and Europe [32,36] and has been running a medical cannabis programme since the late 1990s [37]. In addition, we found that hospital physicians' and GPs' experiences with prescriptions of medical cannabis were associated with their attitudes towards prescription and beliefs in effects. This may indicate that, over time, physicians and patients may become experienced in using medical cannabis, and that it gradually becomes more used and common. However, more studies focusing on such changes over time are needed to investigate, if the number of years during which medical cannabis has been an opportunity is influential on hospital physicians' and GPs' attitudes and beliefs.

Unanswered questions and future research
This review shows a need for studies using stronger data collection methods to obtain larger study populations and reduce selection bias. Moreover, in the literature search, only two qualitative interview studies were identified, which also emphasizes the need for more qualitative studies on this topic to gain a deeper understanding of physicians' attitudes, experiences, clinical practices, and the factors influencing this. It is reasonable to assume that the use of medical cannabis will increase and expand to more indication areas, as it gets legalised in more countries, and more physicians and patients become familiar with it. Continuous research in this area is needed to increase evidence about effect of medical cannabis and keep awareness of the facilitators and barriers to use medical cannabis as well as potentially induced harm.

Conclusions
This review indicates that hospital physicians and GPs in various specialties experience inquiries about medical cannabis from their patients, and to some extent show openness to provide it, although there was a wide gap between studies in terms of willingness to provide. Most hospital physicians' and GPs' experience a lack of knowledge of beneficial effects, adverse effects and of how to advise patients. Hence, hospital physicians' and GPs' experiences with prescribing medical cannabis, and their knowledge as well as their specialty, may be associated with their attitudes towards prescription and beliefs in effect. More research, including larger studies with cohort designs and qualitative studies, is needed to examine facilitators and barriers to physicians' prescribing practices.