Exploring Clinicians’ Determinants of Participation in Medical Cannabis Decision Making with Older Adults Experiencing Arthritis: A Qualitative Descriptive Study
Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, AB, Canada
EPICORE Centre, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, AB, Canada
Division of Rheumatology, Department of Medicine, Faculty of Medicine & Dentistry, University of Alberta, Edmonton, AB, Canada
Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, AB, Canada
Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, AB, Canada
Cheryl A. Sadowski, Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, 3-229 Edmonton Clinic Health Academy, 11405 87 Avenue NW, Edmonton, AB T6G 1C9, Canada; (cherylas@ualberta.ca).Abstract
Background. Arthritis is a common chronic condition among older adults in Canada. Shared decision making (SDM) regarding medical cannabis (MC) use has been anticipated to result in improved health outcomes. However, the practice of SDM for MC remains limited. Few studies addressed what clinicians need to decide with their patients about MC. We explored clinicians’ perspectives on what determines their decisional needs to participate in MC SDM with older adults experiencing arthritis. Methods. In-depth, semi-structured interviews using purposive sampling were conducted online. An interview guide based on the Ottawa Decision Support Framework (ODSF) was used. Recordings were transcribed verbatim, and the data were analyzed deductively and inductively using reflexive thematic analysis. Results. Semi-structured interviews with 12 participants (33% pharmacists, 25% family physicians, 25% rheumatologists, and 17% registered nurses) were conducted. Three main themes were constructed to summarize determinants of clinicians’ needs to be able to participate in MC-related SDM: 1) perception of the decision, 2) perception of others, and 3) decisional conflict. Conclusion. While gaps in evidence remain a critical concern affecting inadequate knowledge and contributing to decision-making needs, findings suggest that a tailored decision support intervention (DSI) would strengthen clinicians’ self-efficacy, clarify role expectations, and attend to the interpersonal dynamics that shape MC SDM. To enhance usability, a DSI should be developed to address clinicians’ decisional needs while remaining responsive to workflow constraints.
Highlights
- Using the Ottawa Decision Support Framework, this study explored clinicians’ decisional needs regarding shared decision making about medical cannabis (MC) use for arthritis.
- Findings highlight gaps that hinder clinicians from engaging in open discussions, leaving many older adults to make MC decisions without professional guidance.
- Addressing clinicians’ decisional needs is essential for developing decision support interventions and policies that ensure safe, informed care.
Arthritis is a common chronic condition among older adults in Canada, with prevalence estimates of 44.3% among those aged 65 y and older (40.7% in men and 46.7% in women), reaching nearly 1 in 2 in this age group.1bibr2-23814683261444057bibr3-23814683261444057–4 Among this population, arthritis is frequently accompanied by debilitating pain, reduced mobility, and associated comorbidities such as anxiety and depression, which together substantially diminish quality of life.5,6 Despite advances in conventional treatments, patients continue to report insufficient symptom relief, leaving them with a persistent burden of pain and distress.7,8
Faced with inadequate symptom control, some older adults turn to alternative options such as medical cannabis (MC), particularly following its legalization in Canada in October 2018.9bibr10-23814683261444057–11 In this study, MC is conceptualized as the therapeutic use of cannabis–based products, including its primary constituent cannabinoids delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD), for the management of health-related symptoms—rather than for recreational purposes—and is typically pursued with or without medical guidance to alleviate pain, improve sleep, or reduce anxiety associated with arthritis.9,12bibr13-23814683261444057bibr14-23814683261444057–15 Legalization and improved access have facilitated cannabis use among older adults with arthritis. Promoted for its potential to alleviate chronic pain, improve sleep, and reduce anxiety, cannabis has attracted substantial attention among those seeking better symptom management, introducing both new opportunities and challenges for patients and health care providers (HCPs).12,13 The decision to use MC for arthritis symptom relief, however, is highly individualized and shaped by personal preferences and values.14,15 This reflects a preference-sensitive decision, where no single treatment option is universally “correct,” and patients’ informed preferences are central to decision making.16,17 Shared decision making (SDM) is particularly essential in such contexts, as it has been associated with improved patient knowledge and greater alignment between patient values and treatment choices, leading to increased satisfaction with care.17bibr18-23814683261444057–19
SDM has been increasingly endorsed as an optimal approach for navigating MC-related decisions15,20 as it involves a collaborative process in which clinicians and patients work together to make decisions that align with evidence and match patients’ values.21,22 SDM requires active engagement, whereby clinicians integrate their clinical expertise with the patient’s lived experiences, expectations, preferences, and values.15,20,23,24 Such a collaborative process ensures that the final decision is not only shaped by clinical evidence and clinician expertise but also congruent with the patient’s unique values and life context—a consideration particularly important in the preference-sensitive context of MC decision making.15,22
Despite the proven benefits of SDM in improving decision quality and reducing decisional conflict for preference-sensitive decisions, implementing SDM for MC decision remains particularly challenging, as clinicians frequently report limited ability to participate in MC discussions with patients.25,26 This situation can amplify patients’ uncertainty and complicate clinical encounters. The Ottawa Decision Support Framework (ODSF) highlights how unmet decisional needs, such as a lack of external resources (e.g., informational support) and internal ones (e.g., self-efficacy), can serve as barriers to participation in SDM. Therefore, understanding decisional needs that influence clinicians’ willingness and ability to engage in these conversations is critical, as these factors directly shape whether and how MC SDM can be implemented.
Decisional needs refer to the deficits that can adversely affect the quality of a patient’s decision (i.e., being informed and matching their values). 27 Identifying these decisional needs among clinicians is essential to developing tailored interventions to support MC SDM.27,28 As such, this study aimed to explore clinicians’ perspectives of what determines their decisional needs to participate in MC-related decision making with older adults living with arthritis. By identifying decisional needs, we will inform the development of an evidence-based, theory-driven, tailored decision support intervention (DSI) that helps clinicians and older adults living with arthritis to better participate together in making quality decisions about using MC.
Methods
Study Design
This qualitative descriptive study29,30 was conducted and reported in alignment with the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist to ensure transparency and rigor. 31 We used ODSF to guide the exploration of decision-making processes, as it is a widely applied framework integrating concepts from social support, cognitive, and psychological theories to inform the development of DSIs. The qualitative descriptive approach was chosen to ensure the data remained close to the participants’ own words (i.e., staying close to the “surface” of the data),32,33 allowing for a detailed description of the perspectives of participants.34bibr35-23814683261444057–36 Participants were eligible for inclusion if they were clinicians who provided care to patients with arthritis and had either received patient inquiries regarding MC or prescribed MC for their patients.
Data Collection
Recruitment of participants took place from July 2024 to December 2024. To ensure broad representation of clinical settings and avoid institutional bias, we recruited participants from diverse practice settings across the province of Alberta rather than a single health care institution. We included clinicians from a range of professional backgrounds, including pharmacists, rheumatologists, family physicians, and nurses, to capture diverse perspectives from clinicians who were likely to encounter MC-related questions in routine practice. In the Canadian health care system, pharmacists are often well integrated into primary care teams and, in Alberta, have an expanded scope of practice that may include prescribing for some conventional medications among authorized pharmacists. Pharmacists commonly contribute to MC-related decision making through patient counseling, screening for drug–drug interactions, harm reduction, and interprofessional collaboration in primary care and community settings. Their inclusion was therefore intended to reflect their decision-support role, given pharmacists’ accessibility within the health care system. 37 We employed purposive sampling to recruit clinicians who were eligible based on the inclusion criteria. This was supplemented by snowball sampling, in which participants identified colleagues who might be eligible.
Recruitment advertisements were mainly distributed through 2 channels: 1) provincial professional newsletters circulated by the Alberta College of Pharmacy and the Alberta Rheumatology Network, reaching practitioners across Alberta and 2) targeted e-mail invitations sent through the researchers’ professional networks and personal connections to family physicians and nurses practicing in arthritis care. This multichannel approach allowed for a wide reach across clinical settings, ensuring the sample was not clustered within a single network or institution. A recruitment flow diagram is available to summarize participant identification, response, and inclusion (Supplementary Figure S1). Data were collected using semi-structured interviews to explore each participant’s perspectives in depth, allowing for probing into individual experiences and clarification of complex issues, which was essential for capturing the nuanced decisional needs relevant to participation in the MC use decision. The interview guide was informed by the ODSF and adapted from the Decisional Needs Assessment Workbook, Jacobsen et al., 28 which is designed to assess the decisional needs of clinicians across diverse health care settings. The guide was reviewed by the research team and iteratively refined during early interviews to enhance clarity (Supplementary File S2). The semi-structured format allowed flexibility consistent with qualitative best practices.
All interviews were conducted via Zoom for 30 to 40 min by the lead researcher (H.A.T.A.), who worked as a graduate research assistant and had previous experience with qualitative research projects, ensuring a consistent audio recording of each session for accuracy and clarity. Interviews were uploaded into Quirkos software and transcribed verbatim. 38 The interviewer had no bias or assumptions about MC use. For accuracy, transcripts were reviewed and revised by the researcher (H.A.T.A.). Additional data sources were notes taken during the interview by the same researcher, memos made during data analysis, and a literature review.39,40
Data Analysis
Data analysis followed a qualitative descriptive approach guided by the ODSF as the primary analytic lens to identify and categorize clinicians’ decisional needs regarding MC SDM.27,29,41 An iterative analytic strategy was used, and analysis was conducted concurrently with data collection to allow constructed insights to inform subsequent interviews.30bibr31-23814683261444057–32 The analytical process proceeded through initial coding, thematic synthesis, and achieving rigor.
The coding process used both deductive and inductive approaches. 42 Deductive coding was informed by the ODSF constructs through operationalizing “decisional needs” as described by Hoefel et al., 27 adapted to the context of MC decision making to identify deficits hindering clinicians’ participation in MC use decisions.27,43 Concurrently, the analysis remained open to participant-driven concepts through inductive coding. The lead researcher (H.A.T.A.) familiarized herself with the data through repeated readings of transcripts and then coded transcripts to capture both semantic and conceptual meanings. 41 Quirkos software was used to organize and code the data. 38 Following initial coding, reflexive thematic analysis was conducted according to Braun and Clarke’s systematic approach. 41 Codes with similar meanings were iteratively grouped into higher-level categories and themes, balancing close adherence to participants’ words with the construction of themes. Themes were refined, reviewed, and named through iterative discussion, with H.A.T.A. and C.A.S. reexamining codes after the preliminary coding phase to refine the thematic construction.
Continuation of recruitment was guided by the principle of information power. 44 Rather than relying on thematic saturation that does not align with reflexive thematic analysis, interviews were discontinued once the data were judged to provide sufficient informational richness to address the study aim. This approach was informed by our focused research question, the high specificity of the clinician sample, the use of the ODSF to guide analysis, and the depth of the interview data.
Achieving Rigor (Trustworthiness)
The completed COREQ checklist is provided to support transparent reporting (Supplementary Table S3). We followed established strategies to ensure rigor (i.e., trustworthiness) by providing a detailed (i.e., thick) description of the findings, describing participants and setting, using probes in the semi-structured interview questions, keeping a document of decision making, and conducting reflexive journaling.45,46 In addition, we used field notes during and after the interviews, memos, and relevant literature reviews on MC decisional needs as multiple data sources (i.e., triangulation). 40 Reflexivity was upheld by consistently documenting decisions, reflections on data collection and analysis, and aligning these reflections with the data during analysis. 47 All findings were presented to the coauthors during several meetings to refine the themes and coding, ensuring the credibility of the findings. Ethics approval was received from the University of Alberta.
Results
Twelve clinicians (33% pharmacists, 25% family physicians, 25% rheumatologists, and 17% registered nurses) participated in the study. A summary of the participants’ characteristics is presented in Table 1. Participants’ ages ranged from 29 to 75 y, and their clinical experience ranged from 1 to 50 y. Three key themes (Figure 1) were constructed to reflect clinicians’ perspectives of what determines their decisional needs to support patients during MC SDM: 1) perception of the decision, 2) perception of others, and 3) decisional conflict. Table 2 provides a summary of themes, subthemes, codes, and representative quotes showing clinicians’ determinants of participation in MC decision making.
| Characteristic | n (%) |
|---|---|
| Gender | |
| Male | 5 (42) |
| Female | 7 (58) |
| Age, y | |
| 25–35 | 1 (8) |
| 36–45 | 6 (50) |
| 46–55 | 2 (17) |
| 56–65 | 0 (0) |
| 66–75 | 2 (17) |
| Not preferring to answer | 1 (8) |
| Medical profession | |
| Rheumatologist | 3 (25) |
| Pharmacist a | 4 (33) |
| Family physician | 3 (25) |
| Nurse b | 2 (17) |
| Year of practice | |
| 1–10 | 3 (25) |
| 11–20 | 5 (42) |
| 21–30 | 2 (17) |
| 31–40 | 0 (0) |
| >40 | 2 (17) |
| Approximate number of cannabis inquiries received per week | |
| 0–1 | 5 (42) |
| 1–2 | 4 (33) |
| 2–3 | 1 (8) |
| 3–4 | 1 (8) |
| −5 | 1 (8) |
| Approximate number of patients seen each week | |
| 0–25 | 2 (17) |
| 26–50 | 6 (50) |
| 51–75 | 1 (8) |
| 75–100 | 0 (0) |
| >100 | 3 (25) |
| Approximate number of patients with arthritis seen each week | |
| 0–25 | 8 (67) |
| 26–50 | 4 (33) |
| How often clinicians are involved in the treatment plan | |
| All the time | 9 (75) |
| Most of the time | 2 (17) |
| Sometimes | 1 (8) |
| Theme | Subtheme | Code | Representative Quote |
|---|---|---|---|
| 1. Perception of cannabis use decision | 1.1. Knowledge and informational support | Inadequate knowledge | I’m not going to recommend a different way of doing things with a product that I don’t really know much about. . . . I don’t know about smoking it, or I don’t know anything. Therefore, I don’t probe that much further with my patients because A, I’m not an expert and B, I don’t understand cannabis grams and all that kind of stuff, and so I’m not going to pretend to understand what it means. (ID no. 8, family physician, 15 y of practice) |
| Adequate knowledge | We do not have much information on cannabis efficacy and safety, with no randomized trials or data to provide. What we need is to provide our patients with information so they can make informed decisions. The problem is that there’s not a lot of information regarding cannabis use within our conditions [arthritis conditions], so I cannot provide much because we lack evidence behind its use in our patient populations. . . . I think the lack of information makes it a challenge on both sides” (ID no. 6, rheumatologist, 20 y of practice) | ||
| Inadequate informational support | I don’t have any information resources right now to provide support regarding cannabis use. (ID no. 2, pharmacist, 26 y of practice) | ||
| 1.2. Expectations | Unclear expectations | Cannabis may lead to worsening mental health, there is potential for dependence, you know, and then what is unclear for me is how it can affect their [older adults’] ability to function and work in society. (ID no. 11, family physician, 2 y of practice) | |
| Negative expectations | If we prescribe it to people who ask for it, there will be an increase in certain emergency attendance because of cannabis . . . traffic accidents or accidents at home because they are high and can easily fall from the stairs. (ID no. 3, rheumatologist, 41 y of practice) | ||
| 1.3. Values | Importance of cannabis | Well, it [cannabis] is important as part of the non-official drugs that patients usually use and possibly influence their health. (ID no. 3, rheumatologist, 41 y of practice) | |
| Safety of cannabis and its placebo effect | One major advantage of cannabis that I value is its safety. Safety is the main thing. . . . Also, it may work as a placebo, and I think it is. . . . If it’s taken carefully, it is pretty harmless, especially the CBD. (ID no. 1, rheumatologist, 50 y of practice) | ||
| Patients value it | When I ask the patient, “Do you feel better?” They say, “Yeah, I feel much better!” I think the placebo effect is strong and it does not change their regular conditions, so it could be of value! (ID no. 8, family physician, 15 y of practice) | ||
| Value of cannabis discussion | It [cannabis] is very important, just like any other medication that my patients are on. I want to be able to provide counseling about cannabis comfortably.” (ID no. 9, family physician, 15 y of practice) | ||
| No value of cannabis discussion | It [cannabis decision making] is not an important decision, to be honest. I’m doing it as a favor to you. I think this is such a low-hanging fruit, and I’m really not that clinically interested in this. (ID no. 8, family physician, 15 y of practice) | ||
| 2. Perceptions of others | 2.1. Patients | Disconnection | I feel mostly people feel unsupported . . . they just don’t really know how to go about doing it [using cannabis] and then when they come to me, and I don’t have a lot of information, that doesn’t help them feel more supportive. (ID no. 4, pharmacist, 15 y of practice) |
| Unsatisfactory quality of information brought by patients | I think patients lack evidence-based information and may have confusion from other information overload. (ID no. 4, pharmacist, 15 y of practice) | ||
| Variable experiences of patients regarding cannabis decision making | The cannabis decision is difficult for some patients because they’ve had other clinicians who haven’t been supportive, or again, the cost is a big one. They feel a little bit nervous to spend a lot of money and not know if it will help them. (ID no. 2, pharmacist, 26 y of practice) | ||
| Patients are open to receiving information from clinicians as they trust them | I know they [patients] trust their doctors and they are uncomfortable about the information not being provided through proper medical channels. (ID no. 9, family physician, 15 y of practice) | ||
| Cannabis is a frequent topic raised by patients | Patients usually ask me, what if I can take cannabis? Which cannabis is better, the oil versus the THC? I think the THC is better, isn’t it? This is how they mostly ask about it. (ID no. 5, pharmacist, 9 y of practice) | ||
| Discussions from the patient’s side occur after they have already made the decision | When they [patients] see me, they’ve usually already decided whether they want to take it. They just want my approval! I feel like patients have already made up their minds about cannabis before they come to me . . . they’re not really looking for my input, just my approval. (ID no. 2, pharmacist, 26 y of practice) | ||
| Patients come with fixed beliefs | “They are already in a frame of mind that is positive towards it. (ID no. 1, rheumatologist, 50 y of practice) | ||
| Inquiries were perceived less as requests for decision making and more as efforts to know how to implement the decision | It’s not phrased as a question, so either they say, well, I thought I’d just try smoking weed at bedtime to see if that helped me, but it didn’t. . . . They’ll either say, hey doc, I’m here because of anxiety or I’m here because of pain or I’m here because I can’t sleep. I tried melatonin, but it didn’t work. I tried taking Advil™, but it didn’t work. So I was like, I’ll just try some weed. (ID no. 11, family physician, 2 y of practice) | ||
| Seeking approval | They just kind of, like, lob the idea out. And then wait for some response indicating that I am okay with it [cannabis]. (ID no. 11, 2 y of practice) | ||
| 2.2. Other health care providers’ roles | Need for increased collegial support and collaboration | To be able to practice cannabis shared decision, that needs to be part of a bigger movement involving others. (ID no. 10, pharmacist, 12 y of practice) | |
| Defining whose role it is to discuss cannabis | A pharmacist is trained to deliver what we need to deliver for cannabis decision-making. (ID no. 3, rheumatologist, 41 y of practice) | ||
| Concerns about the appropriateness of nonphysician involvement | It has to be from doctor to patient only. I would not take responsibility. I don’t want to take the decision. . . . The doctor has to supply that [cannabis decision support], not me. (ID no. 5, pharmacist, 9 y of practice) | ||
| 2.3. Patients’ social circles | Families of patients are involved in the decision | Mostly family is involved in the decision, yeah. They do provide support. They do give the information and go from there. (ID no. 9, family physician, 15 y of practice) | |
| Family members might pressure the patient to use cannabis | It’s often their kids or relatives who are pushing towards using cannabis. (ID no. 3, rheumatologist, 41 y of practice) | ||
| The preference is that the decision discussion involves the clinicians and the patient only | Well, if family members are going to discuss it, I prefer it to be between the patient and me only. (ID no. 3, rheumatologist, 41 y of practice) | ||
| 3. Decisional conflict | 3.1. Uncertainty regarding evidence and outcomes | Clinicians’ feeling discomfort toward patients’ inquiries | I don’t feel comfortable responding to cannabis, so I usually defer or try to do additional research. (ID no. 7, registered nurse, 1 y of practice) |
| Perceived poor quality of supporting evidence | Currently available cannabis products that patients use medically are unregulated, as it’s not a medication, then patients are not receiving the information from the health care providers, and health care providers are not actively seeking to give that information either, as there is a lack of good evidence. That makes it a challenge on both sides. (ID no. 6, rheumatologist, 20 y of practice) | ||
| 3.2. Role in MC use decision support | Diminished desire to be involved | I do not want to be involved. I think my answer to any inquiry about cannabis would be that I don’t know for certain what sort of evidence there is around cannabis for arthritis, and that I haven’t heard of anything really robust yet. (ID no. 11, family physician, 2 y of practice) | |
| Desire to be involved | I don’t have any background knowledge about it right now, so I would like to do more research and get more information. That way, I can be involved in helping my patient find a medication that will truly support what they’re going through. (ID no. 7, nurse, 1 y of practice) | ||
| Involvement as passive supporters | I’m really just there to support their decision. . . . I’m certainly in support of it [cannabis use] if they want to take it. (ID no. 2, pharmacist, 26 y of practice) | ||
| Maintaining a neutral stance | I will not discourage the patient from using it; I just want to do more research so I can support them with accurate information. I try to understand what they know or have heard from others, then let them know I need to learn more before guiding them further. (ID no. 7, nurse, 1 y of practice) | ||
| Clinicians expressing ambivalence, discussing cannabis with cautious neutrality | I would react ambivalently in the sense that I would never say, yes, it’s a good idea. I might say to them, well, people have told me that their fatigue or their pain is helped by it. I think if you take this type of CBD, for example, there’s no downside. If you want to try it, try it. But frankly, I’m not particularly recommending it. (ID no. 1, rheumatologist, 50 y of practice) | ||
| Concerns about their self-efficacy | Even if I get some more information about dosage forms available, I still do not feel confident enough to prescribe it. I just can’t. (ID no. 2, pharmacist, 26 y of practice) | ||
| Limited time | Now I’m going to have to tell you that I personally am pretty okay to discuss it as long as the time permits. . . . But when it comes during the busy time of the schedule, I am not comfortable at all. As human beings, every one of us has biases. I can actually go towards the bias more for a quick answer [if I don’t have time] rather than providing a professional, balanced, elaborated answer. (ID no. 5, pharmacist, 9 y of practice) | ||
| Heavy workload | I wouldn’t want to be involved in going through it with the patient . . . it’s time-consuming with my already heavy workload. (ID no. 1, rheumatologist, 50 y of practice) |
Perception of Cannabis Use Decision
Knowledge and informational support
Knowledge and informational support refer to the extent to which individuals are aware of, understand, and have access to essential, accurate, and relevant information needed to make an informed decision. Deficits in this construct reflect gaps in access to information required to support participation in SDM. 27 Participants consistently emphasized that insufficient knowledge served as a primary barrier to engaging in MC-related decision making with patients. All participants acknowledged that their current level of knowledge about MC was inadequate, impeding their ability to contribute meaningfully to MC discussions: “They [patients]’re looking for more answers than I can give! . . . I don’t have sufficient knowledge to answer any cannabis question right away” (ID no. 10, pharmacist, 12 y of practice). The importance of being adequately informed emerged across interviews: “It’s important to know enough about cannabis, so that I get comfortable with prescribing it to my patients” (ID no. 9, family physician, 15 y of practice). Despite recognizing the importance of their role in SDM, some clinicians expressed difficulty in gaining knowledge and providing MC guidance due to inadequate informational support caused by the lack of evidence, unlike other arthritis medications supported by evidence:I’m interested in cannabis SDM as it is my role. But the problem here is that there’s not a lot of information regarding cannabis use within our conditions [arthritis conditions]. I cannot provide a lot of information about it because we do not have evidence like for other treatments. (ID no. 10, pharmacist, 12 y of practice)
Participants frequently expressed frustrations with the absence of robust research, particularly randomized controlled trials: “Most evidence on harms is observational, and while helpful, it’s not the same as RCT [randomized controlled trial] data” (ID no. 8, family physician, 15 y of practice), which made it challenging to provide patients with evidence-based guidance. In addition, the absence of clear clinical guidelines further complicated clinicians’ ability to support patients: “There are no clear-cut guidelines . . . what’s the best dosage form? What is the recommended dosing schedule?!” (ID no. 7, nurse, 1 y of practice).
Expectations
Expectations refer to individuals’ perceptions about the probability of potential outcomes of a decision, including benefits, harms, and other consequences. Expectations are considered unclear when individuals lack awareness of outcome probabilities. 27 In this study, lack of knowledge about benefits and harms appeared to contribute to clinicians’ difficulty in forming clear expectations about cannabis therapeutic potential. “I am not sure what to expect! Am I just creating a new substance use disorder [by prescribing MC] and thinking I’m avoiding an opioid use disorder?” (ID no. 11, family physician, 2 y of practice). While some expressed concerns expecting that MC might replace conventional arthritis therapies, which would worsen arthritis, “I worry sometimes that people will stop taking medications that are actually treating their disease over their ideas that cannabis is a miracle drug, which a lot of people do seem to have that thought” (ID no. 2, pharmacist, 26 y of practice). Others held negative expectations about MC adverse effects: “It [cannabis] can be very badly tolerated if it is not what we think it is” (ID no. 3, rheumatologist, 41 y of practice). Thus, unclear expectations further complicated their willingness to engage in SDM.
Values
Values refer to the importance individuals assign to the features and outcomes of a decision, including benefits, risks, and uncertain outcomes. Values are unclear when individuals have difficulty identifying the features that are important to them. 27 Clinicians expressed varied and sometimes conflicting views regarding the value of MC as a treatment option. While some participants regarded it as potentially beneficial, particularly in managing conditions such as insomnia in comparison with traditional medications,I actually think it [cannabis] is a good option, particularly CBD, for sleep over all of the available choices. . . . For pain relief, I usually like to suggest other options, but if there aren’t any, then I think the topical cream is worth a try. (ID no. 2, pharmacist, 26 y of practice)
Others expressed strong skepticism, dismissing MC as a “false hope” and underscoring concerns about overestimating its therapeutic potential: “Don’t give them [patients] false hope about cannabis. Simple. This will be just another compound going into their body. Nothing more than that” (ID no. 5, pharmacist, 9 y of practice). These divergent views extended to the perceived value of discussing MC with patients. For some, MC-related conversations with their patients were seen as activities that were not currently affecting their practice but could be time-consuming: “Providing or not providing cannabis information to the patient does not at all impact my practice. It takes a few moments of time, though” (ID no. 1, rheumatologist, 50 y of practice). In contrast, others recognized these discussions as meaningful opportunities to build rapport and enhance credibility. “I think it [cannabis discussion with patients] adds to the credibility and the rapport you [as a clinician] have with your patients” (ID no. 10, pharmacist, 12 y of practice). This highlights the importance of engaging in MC-related conversations. Overall, there is a variation among clinicians in valuing MC as an option and consequently valuing discussions with patients involving it, which influences their willingness to participate in MC SDM.
Perception of Others
Patients
Clinicians often described a sense of disconnect between themselves and their patients due to a gap in approaches to MC decision making. “Cannabis users are on Mars, and clinicians are on Venus! They’re not even in the same stratosphere. Where we’re approaching it, and where they [patients]’re approaching it, there are 2 nonoverlapping Venn diagram.” (ID no. 8, family physician, 15 y of practice). Another added, “I’m not within their [patients’] shoes, so I can’t make comments in terms of what they feel or do regarding cannabis” (ID no. 6, rheumatologist, 20 y of practice). This gap was attributed in part to the quality of information patients brought from nonclinical (e.g., Google) or anecdotal sources (e.g., friends). “It’s hard to have a conversation about cannabis with patients because they come in with so much information already from Google or their friends, and I don’t know how much of it is actually accurate” (ID no. 10, pharmacist, 12 y of practice).
Clinicians perceived that patients experienced varying degrees of ease when making MC-related decisions. Cannabis decisions appeared to be easily made by some patients. “Accessibility of cannabis makes it a very easy decision for the patients” (ID no. 8, family physician). However, clinicians indicated that other patients may approach the decision with difficulty due to lack of clinicians’ support and the high cost of MC: “Cannabis decision is difficult for patients because other clinicians haven’t been supportive or the cost is a big one. They feel nervous to spend a lot of money and not know if it will help them” (ID no. 2, pharmacist, 26 y of practice).
Despite these variations among patients, clinicians acknowledged that patients often placed considerable trust in them and were open to receiving their input. “They do put a lot of trust in me, so I know they would be very open to getting information about cannabis from me” (ID no. 9, family physician, 15 y of practice). Cannabis was a frequent topic raised by patients: “Cannabis use is quite common, so it does come up often during visits” (ID no. 1, rheumatologist, 50 y of practice). The topic was often raised indirectly through personal stories or exploratory questions seeking approval rather than explicit clinical inquiries. “They just kind of, like, lob the idea out. And then wait for some response indicating that I am okay with it [MC]” (ID no. 11, family physician, 2 y of practice). These discussions typically occurred after patients had already decided to use MC often grounded in beliefs about its benefits, with clinicians interpreting such interactions as efforts to seek validation. “They come to see me with fixed beliefs about its benefits because they know somebody who used it and did better. So, they give it a go, then they talk to me!” (ID no. 8, family physician, 15 y of practice). In other instances, inquiries were perceived less as requests for decision making and more as efforts to know how to implement the decision. “When they come to see me, I feel like they’ve already made the decision to try cannabis, but they just don’t really know how to go about doing it” (ID no. 4, pharmacist, 15 y of practice). In summary, clinicians perceived a disconnect with patients who, guided by fixed beliefs and non–evidence-based information, had often decided to use MC and sought validation rather than open discussion.
Other HCPs’ roles
Clinicians described varied perceptions of the roles of other HCPs in MC-related decision making, including the need for greater collegial support, differing views on which professionals should be involved, and uncertainty regarding where responsibility for MC deliberation should reside. Some participants indicated that their colleagues are not interested in MC decision making. “I think many of our clinicians are not very interested in the discussion about cannabis decision making because patients are using it anyway” (ID no. 8, family physician, 15 y of practice). Many participants emphasized the need for increased collegial support and collaboration:I think cannabis discussion with patients could be easier for me if others in my clinic were more knowledgeable and supportive of it. They often have a negative idea about it when, in many cases, it can actually be helpful for patients. (ID no. 2, pharmacist, 26 y of practice)
Several participants expressed a preference for patients to consult other HCPs with specialized training in this area. “Cannabis decision making wouldn’t be part of my clinic . . . that would be another HCP who’s actually prescribing it as they have a special training for that” (ID no. 2, pharmacist, 26 y of practice). Others added that “I know there are clinics that just specialize in cannabis prescription. So that’s an ideal place if a patient is interested” (ID no. 12, nurse, 15 y of practice). For defining whose role it is to discuss MC, some clinicians acknowledged that pharmacists could play a role. “Whether it’s the pharmacist or the people at the dispensaries, they should definitely be involved in cannabis decisions” (ID no. 11, family physician, 2 y of practice). However, others expressed concerns about the appropriateness of nonphysician involvement, suggesting that MC decisions should remain within the domain of physicians or other specialists.
The cannabis decision should be made from a very top level. I would say from a doctor or a specialist. . . . Any cannabis discussion should be between the patient and their physician rather than the patient and the pharmacist. Let’s have the pharmacist be out of this. . . . Involving more professionals in that decision [cannabis use decision], even some nurse practitioners would probably have better outcomes. (ID no. 5, pharmacist, 9 y of practice)
Overall, clinicians expressed a need for greater support from their colleagues, as each profession appeared to shift the responsibility for MC deliberation onto others.
Decisional Conflict
Decisional conflict is defined as the uncertainty about which course of action to take when options involve risk or challenge to one’s personal values, often experienced when individuals feel uninformed, unclear about their values, or unsupported in making choices, and it is usually manifested as verbalized uncertainty. 27 From the clinicians’ perspective, this conflict emerged in relation to the MC use decision in itself and whether to actively participate in MC SDM.
Uncertainty regarding evidence and outcomes
Clinicians expressed decisional conflict regarding MC use reflected as discomfort, often linked to a lack of knowledge leading to unclear expectations: “Not knowing what kind of side effect it [cannabis] could have . . . that makes me very uncomfortable if I am asked about it” (ID no. 9, family physician, 15 y of practice). Others described more intense feelings of fear: “When I am asked a question about cannabis, it’s a fear that I feel because I have to answer their question very well, while I can’t” (ID no. 4, pharmacist, 15 y of practice). Decisional conflict seemed to be often rooted in the uncertainty about its efficacy, leading to just telling patients about the lack of evidence and agreeing with them. “I’m not certain what evidence is around cannabis for arthritis. I tell my patients, ‘I’m not sure if it [cannabis] will help your fatigue, but if fatigue is a problem and you want to try it, give it a try!’” (ID no. 1, rheumatologist, 50 y of practice). Even if evidence exists, its quality was questionable, which compounded uncertainty. “If there is any evidence, it’s really not done well or not long-term, therefore I am not sure if cannabis could be useful or harmful” (ID no. 8, family physician, 15 y of practice). Together, these accounts illustrate that clinicians’ decisional conflict was primarily driven by uncertainty regarding its clinical benefit, with concerns over the absence or perceived poor quality of supporting evidence.
Role in MC use decision support
For many clinicians, decisional conflict regarding MC contributed to a diminished desire to engage in MC SDM. For a smaller number of clinicians, the desire to be involved was evident; however, barriers such as nonstandardized MC products, limited peer support, inadequate training, and time constraints limited them. Clinicians perceived inconsistent product quality and lack of standardization as barriers to engaging in MC SDM. “I don’t know about the quality and standards of cannabis products. We need pharmaceutical-grade products so that we can support our patients” (ID no. 3, rheumatologist, 41 y of practice). Several participants framed their involvement as passive supporters agreeing with the patient’s decision without offering direct guidance. “My role is providing support and advice . . . but the decision is totally theirs . . . I do not want to be involved” (ID no. 1, rheumatologist, 50 y of practice). Others acknowledged patient use without endorsing it, maintaining a neutral stance: “I’m not advising patients to take it [cannabis], but I’m certainly not neglecting it either” (ID no. 3, rheumatologist, 41 y of practice). Some explicitly avoided MC discussions, preferring to refer patients to other providers. “When they ask about it [cannabis], I usually just send them back to their doctor. . . . I would prefer a referral over providing information. . . . I don’t feel it’s my role to prescribe cannabis or take the lead on its use” (ID no. 5, pharmacist, 9 y of practice). Participants expressed concerns about their self-efficacy in supporting patients with MC use decisions, citing a lack of proper training and uncertainty about the credibility of available education. “We’re not ready for it [cannabis use SDM]. We were not properly trained for it. Plus, whoever will be giving us the training, we don’t know how biased they are to push that into the market” (ID no. 5, pharmacist, 9 y of practice). One participant indicated a perceived undefined decisional role as legalization and infrequent clinical exposure contributed to disengagement from MC-related care.
Patients do not ask me often about cannabis because it’s available everywhere for anyone . . . if it wasn’t legal, then they probably would ask me for medical authorization . . . as I do not encounter it every day in my practice, I wouldn’t be proficient at it. (ID no. 12, nurse, 15 y of practice)
Some clinicians indicated that limited time is a barrier to learn about MC.
I just don’t have time to learn about cannabis. I have a very busy job . . . I work about 45 h a week. I just don’t have time to take on any extra things like that [supporting patients about using cannabis]. I already have sort of niche pieces within my clinic that I know enough that I can talk about them. (ID no. 2, pharmacist, 26 y of practice)
Others emphasized the difficulty of fitting MC conversations into their already heavy workload.
Unfortunately, right now in primary care, nobody’s got time to do the proper job. We all struggle with not having enough time to do what we want to do in a day. So I don’t think I would have time to go through a cannabis discussion. (ID no. 8, family physician, 15 y of practice)
Overall, clinicians often positioned themselves on the margins of MC-related decisions, describing their perceived involvement as lacking significance, rather than as meaningful active partnerships with their patients. An interpretive synthesis illustrating the relationships and interplay among the constructed themes is presented in Figure 2 and further elaborated in the “Discussion” section.
Discussion
This qualitative study explored clinicians’ perspectives on what determines their decisional needs to participate in MC-related decision making with older adults living with arthritis. In the context of growing patient interest in using MC for its therapeutic effect on arthritis symptom management, where policy, evidence, and clinical practice are not aligned, clinicians are increasingly approached by patients to engage in MC discussions.15,25,26 Within this context, the complexity of these MC-related discussions is further compounded by an evolving yet still uncertain evidence base relevant to arthritis symptom management. 48 Recent evidence documented expanding research activity on cannabinoids for chronic pain, including arthritis, but consistently highlighted variability in effects and a persistent lack of high-quality data to inform long-term efficacy and safety, particularly among older populations.48bibr49-23814683261444057bibr50-23814683261444057bibr51-23814683261444057bibr52-23814683261444057bibr53-23814683261444057bibr54-23814683261444057–55 This uncertainty is reflected in key clinical challenges reported in the literature, including the absence of standardized dosing units that complicates prescribing and counseling practices, 56 difficulties related to medication substitution when patients transition from conventional therapies to MC, 57 and ongoing questions regarding optimal dosing, safety, and medico-legal considerations.53bibr54-23814683261444057bibr55-23814683261444057bibr56-23814683261444057–57 While professional rheumatology organizations have issued pragmatic position statements to guide clinicians navigating this landscape, 58 these recommendations are applied within a context where definitive clinical guidance remains limited. 59 Collectively, this growing yet inconclusive body of evidence underscores how MC-related decision making for arthritis is shaped not only by patient interest but also by clinicians’ decisional needs and barriers to support such a preference-sensitive decisions. 59
Despite this expanding MC literature, much of the existing research continues to focus primarily on patients’ perceptions of MC use, with comparatively less attention given to exploring clinicians’ decisional needs.15,25,26,60,61 Understanding such needs is essential to inform the design of meaningful interventions that facilitate SDM in today’s busy clinical settings.27,28 Key themes identified in this study were synthesized and situated within the existing literature,62bibr63-23814683261444057–64 highlighting decisional conflict as a central construct that influenced clinicians’ behavior and perceived roles and was shaped by underlying decisional needs.
Building on this synthesis, in the present study, decisional conflict was constructed as a key integrative decisional need across the data, contributing to the extent of clinicians’ engagement in MC SDM. Decisional conflict was the cumulative outcome of multiple, intersecting decisional needs and barriers that seemed to interact, leading to the final outcome of decisional conflict. Those decisional needs mainly emerged from limited external resources (informational support), leading to inadequate knowledge, unclear expectations, and diverse interpretations of value, both in terms of MC therapeutic effects (harms and benefits) and the value of initiating discussions about it with patients. Coupled with decisional barriers, these decisional needs contribute to decisional conflict, resulting in limited internal resources such as self-efficacy, as shown in Figure 2. This finding aligns with prior research describing decisional conflict as a multidimensional construct arising from unmet decisional needs across informational, values clarification, and support domains in complex clinical contexts.62bibr63-23814683261444057bibr64-23814683261444057bibr65-23814683261444057bibr66-23814683261444057bibr67-23814683261444057–68 Consistent with the ODSF,27,44 decisional conflict in this context may therefore serve not only as an outcome but also as a main contributing factor to other unmet decisional needs that shape clinician engagement in MC SDM and can be addressed through evidence-based DSI development.
Among the barriers reported by participants were the uncertainties regarding product quality and the lack of standardization that constrained the clinicians’ ability to fulfill their decisional support role. Without reliable, pharmaceutical-grade products, clinicians perceived that limitations in providing evidence-based guidance were amplified, impeding SDM. These concerns reflected gaps in knowledge and support, consistent with the decisional needs constructs of the ODSF27,65 and aligned with existing literature highlighting challenges related to the regulation, quality assurance, and clinical use of MC.58,69,70 Furthermore, the interplay between decisional conflict and unmet decisional needs and barriers appeared to negatively affect clinicians’ self-efficacy to engage in MC-related SDM, as participants indicated discomfort when responding to patient inquiries. This observation is in agreement with prior work documenting clinicians’ limited confidence in their MC knowledge and its implications for clinical engagement. 26
As such, the decisional needs and barriers specific to the MC context contributed to decisional conflict, which appeared to negatively influence clinicians’ self-efficacy and limit their clarity regarding their role in MC SDM. Decisional conflict, when coupled with clinicians’ perceptions of patients and their social networks/families, contributed to interpersonal tensions that further complicated MC-related discussions. Clinicians viewed the combination of informal information seeking, indirect communication due to social stigma, and postdecision engagement as limiting the potential for timely, collaborative MC SDM. These dynamics were consistent with prior literature describing attitudes, perceived risks, and informal information pathways surrounding MC use. 61 Participants reported that patients’ social networks, including family members and peers, often played a prominent role in encouraging MC use, placing further pressure on both patients and clinicians. Tension appeared to arise when clinicians’ unmet decisional needs, such as limited knowledge and self-efficacy, intersected with patients’ implicit demands, pressure, and expectations for validation. This finding was consistent with other research documenting patients’ perceptions and preferences regarding MC use9,12,13,15,20,24,60,61 and clinicians’ beliefs, confidence, and practice related to MC.71bibr72-23814683261444057bibr73-23814683261444057bibr74-23814683261444057–75 Together, these interpersonal dynamics illustrated how clinicians’ decisional conflict misaligned with patients’ expectations, contributing to a relational strain, especially when clinicians felt that patients expected them to support a course of action they did not fully understand or agree with. That tension disrupted communication and hindered the practice of MC SDM, mirroring broader communication barriers identified in other research.76,77
Coupling all the previously mentioned decisional needs and barriers shaping decision conflict with clinicians’ perceptions of patients as implicitly pressuring them to approve MC use, clinicians expressed a difficult decisional role. According to the ODSF, a difficult decisional role is a decisional need that arises when there is a mismatch between one’s preferred and actual involvement in decision making.27,65 Clinicians in this study often indicated their genuine desire to support their patients; however, they also described discomfort when responding to patient inquiries and, at times, avoidance of responsibility. Such indications emphasized a perceived difficult decisional role, a perception previously reported in similar clinical contexts characterized by high uncertainty.78bibr79-23814683261444057–80
Similarly, prior research indicated that deficits in external resources (e.g., limited access to credible MC information, time constraints, and lack of peer support) and internal resources (e.g., diminished self-efficacy), together with perceived pressure from patients, contributed to decisional conflict, which in turn compounded clinicians’ decisional role distress, leaving them uncertain about how to engage in MC-related SDM even when motivated to do so.27,28,56bibr57-23814683261444057bibr58-23814683261444057–59,65,66,68,81bibr82-23814683261444057–83 Consequently, as clinicians indicated that their decisional needs and barriers were not addressed, they also expressed a tendency to defer responsibility to other providers, such as cannabis specialists, whom they perceived as better positioned to lead MC SDM. This preference aligned with other research reporting about clinician decisional conflict and referral practice.26,59,76,84 Taken together, these findings illustrated how decisional conflict in the present study translated into limited self-efficacy, difficult decisional roles that prompted responsibility shifting, and relational tension that, in turn, constrained MC SDM, consistent with broader evidence on clinician uncertainty and practice.25,73,75
By illuminating these decisional needs and barriers, this study underscored the critical need for DSIs that directly address decisional conflict to strengthen clinician decisional role clarity and ultimately facilitate MC SDM. This descriptive qualitative study is strengthened by its documentation of inclusion of a diverse group of clinicians—pharmacists, rheumatologists, nurses, and family physicians—from various health care settings. Such documentation of demographic and professional diversity has been reported to enhance the transferability of the findings across clinical contexts.45,85,86 Conducting interviews via Zoom further enabled geographic representation from across a region. However, given the sensitive nature of MC-related decision making, participants may have moderated their responses due to stigma, historical illegality, and concerns about professional liability, challenges commonly reported in research on MC.72,87bibr88-23814683261444057–89 Social desirability bias may also have influenced responses, with clinicians possibly presenting more professionally acceptable views as reported in other research. 90 Despite these limitations, the study provided useful insights into clinicians’ perceived decisional needs and highlighted the need for further research and policy attention to support MC-related initiatives that facilitate MC SDM in arthritis care for older adults.
Conclusion
Our findings underscored the impact of clinicians’ perceptions of MC decisions, others involved in the decision, and decisional conflicts on shaping their engagement in MC SDM. Clinicians’ limited engagement in MC SDM appeared to reflect the convergence of interacting decisional needs and barriers that contributed to decisional conflict, which, in turn, could contribute to undermined self-efficacy, strained clinician–patient relationships, and the difficult decisional roles of clinicians. Unlike many other pharmacologic decisions in rheumatology, MC-related discussions seemed to be particularly complex due to shifting legal frameworks, inconsistent product quality, lack of standardization, social stigma, and the scarcity of high-quality clinical evidence as reported in prior research.58,60,91 These distinguishing features differentiate MC from conventional treatment options and underscore the need for DSIs that are purposefully designed to address these challenges to facilitate SDM. While gaps in evidence remain a critical concern, these findings suggested that effective DSIs should also strengthen clinicians’ self-efficacy, attend to the interpersonal dynamics that shape MC consultation, and clarify decisional roles. To be implementable in daily clinical practice, such interventions should be developed to address clinicians’ decisional needs while remaining responsive to workflow constraints that affect engagement in SDM.