Patient Perspectives on Opioid Risk Discussions in Primary Care

Abstract

Purpose

Both patients and clinicians have described discussions of potential opioid risks as challenging. This study’s goal was to understand patient perspectives on discussing opioid risks with primary care clinicians (PCCs).

Methods

Patients identified to be at elevated risk for problems with opioids (ie, opioid use disorder [OUD] diagnosis, taking a medication for OUD, or having ≥3 opioid prescriptions in the last year) were recruited from an integrated, Upper Midwest health system to participate in semi-structured qualitative interviews. Interview questions aimed to better understand patient views on conversations about opioid risks with PCCs and perceptions of OUD screening and treatment in primary care. Interviews were analyzed using an inductive thematic analysis approach.

Results

A total of 20 patients participated (mean age: 53.5 years; 65% male). Six themes emerged: 1) archetypes of patient relationships with opioids (long-term opioid use, acute opioid use, OUD in treatment, OUD no treatment) require different approaches in discussing opioid risks; 2) patients may develop their own archetypes about PCCs and opioids; 3) these archetypes may help guide how conversations about opioids are conducted (eg, PCC demeanor, terminology); 4) most patients believe that primary care is an appropriate setting for opioid risk discussions; 5) patients may have limited awareness of the availability and value of overdose rescue medications; and 6) handouts are more acceptable if perceived to come from the PCC’s assessment instead of a computer.

Conclusions

Results suggest that patients generally perceive discussing opioid risks with PCCs acceptable. PCCs should tailor opioid risk conversations to patients’ specific situations and needs.

Approximately 68,000 Americans died from opioid-related overdoses in 2020, a 37% increase from 2019.¹ The most effective treatments for opioid use disorder (OUD) are OUD medications (MOUDs) — methadone, buprenorphine/naloxone, and extended-release naltrexone — which reduce opioid cravings, relapse, and overdose.² However, only about 1 in 5 people with OUD receive such treatment.³

Increasing MOUD access for patients with OUD in outpatient settings is one approach to help end the opioid epidemic. Specifically, buprenorphine/naloxone can be prescribed in office-based outpatient settings by clinicians with a Drug Addiction Treatment Act of 2000 (DATA-2000) waiver.⁴ People with substance use disorders may be more willing to seek treatment in primary care settings than at specialty drug or alcohol treatment centers, likely because they perceive less stigma in seeking care in primary care settings.⁵ However, relatively few primary care clinicians (PCCs) have obtained waivers, and of those waivered, many do not prescribe buprenorphine/naloxone.⁴ Among both waivered and nonwaivered clinicians, lack of time, training, resources, and confidence are barriers to prescribing MOUDs.⁶ Encouraging PCCs to address opioid risks could increase treatment engagement, as recent data suggest that people with substance use disorders who report being screened by a health care provider are 3–4 times more likely to receive treatment than those who are not screened.⁷

To address barriers to OUD care in primary care, our team developed a shared decision-making (SDM) tool embedded in the electronic health record (EHR) that alerts PCCs when a patient may be at high risk for OUD or overdose and offers support for screening, diagnosing, and treatment (trial information available at https://clinicaltrials.gov/ct2/show/NCT04198428). The primary goal of this tool is to prompt discussions about opioid risks and to promote appropriate screening and treatment. Although PCCs acknowledge their responsibility in discussing opioid risks with their patients,⁸ they also find these conversations challenging.⁹ Some PCCs report negative attitudes toward people on MOUDs (eg, they are difficult, argumentative, manipulative, or undesirable in clinical settings)⁶ and may avoid having conversations with patients about OUD treatment when possible.¹⁰ Clinicians vary in how they communicate opioid risks to patients, but the interpretations and impact of these variations with patients in unclear.^11,12

On the patient side, some literature suggests that people who use opioids to manage chronic pain may perceive discussions about opioid risks with PCCs as stigmatizing.¹³ Patients may not understand that having a substance use disorder is not a moral choice, and they may be concerned that PCCs will accuse them of inappropriate behavior rather than approaching out of concern for their well-being. Patients also report lower satisfaction of and more difficulty with encounters where clinicians and patients disagree on an opioid management plan.¹⁴ Thus, PCCs and patients may benefit in working together to minimize opioid risks, but little is known about patients’ preferences for language and settings for these conversations.

This study aimed to learn patient perspectives on how our SDM tool may be received and to gain an understanding of how PCCs might best approach opioid risk discussions in primary care. We posed the following research questions: 1) How do conversations about opioid risks with PCCs typically go? 2) How do patients feel about being screened or treated for OUD in primary care? 3) How do patients perceive different messages about opioid risk? and 4) How do patients feel about receiving a handout stating their risk? To answer these questions, we conducted a series of semi-structured qualitative interviews with patients whose clinical characteristics would prompt the SDM tool in the EHR.

METHODS

Study Setting and Design

This qualitative, preplanned substudy was conducted during the design phase of a larger clinical trial examining the effect of an EHR-linked SDM tool to help PCCs screen, diagnose, and treat patients with OUD. The intervention was implemented in a large multispecialty care system with 600 PCCs practicing in 52 clinics in the metropolitan Minneapolis/St. Paul area of Minnesota. The study was reviewed, approved, and monitored by the HealthPartners institutional review board (study A18-345).

Participants

Eligible study patients were identified for recruitment by 8 PCCs interviewed in the design phase⁸ for optimizing the implementation and use of the SDM tool from the larger clinical trial (https://clinicaltrials.gov/ct2/show/NCT04198428). The research team pulled a list of eligible patients from the EHR who would be targeted for the SDM tool and 1) were at least 18 years old; 2) had a primary care encounter with the identifying PCC within 6 months of the study invitation (between March and October 2019); and 3) had an OUD diagnosis, were taking a MOUD, or had ≥3 opioid prescriptions in the past year. Patients who were prisoners or currently hospitalized were ineligible. PCCs reviewed lists of patients who met inclusion criteria and identified candidates who would be inappropriate to contact. While PCCs were not required to disclose reasoning for discretionary exclusion, possible scenarios include declining physical health of patient, complex psychosocial situations, or the clinician not having a strong relationship with the patient (eg, had only seen once or twice).

Research assistants mailed invitations to 59 patients and followed up by phone to determine patient interest and eligibility. Of the 59 patients, 37 (63%) were reached via phone. Of these, 23 (62%) were scheduled for interviews, 12 (32%) declined participation, and 2 (5%) were ineligible. In all 20 patients completed the interviews and were included in the analysis. Data saturation was reached after the 20th interview, and no additional interviews were scheduled.¹⁵

Procedure

Semi-structured interviews were conducted, recorded, and transcribed between December 2019 and March 2020 in-person or by phone with a research project manager trained in qualitative interviewing. Interviews lasted 20–30 minutes and followed a guide with 7 stem questions about patients’ 1) relationships with their PCC; 2) experiences taking opioids; 3) past discussions and perceived importance of opioid risks with their PCCs; 4) discussing opioids with their PCCs at care encounters for other complaints; 5) preferred OUD terminology (eg, opioid problems, dependence, addiction, misuse); 6) reactions to targeted messaging and SDM handouts; and 7) advice to PCCs for discussing opioid risks.

The research team debriefed after every 2–3 interviews to identify important highlights relevant to the ongoing SDM tool design, summarize immediate impressions, and inform subsequent interviews. During the debrief meetings, the team determined that data saturation was reached after no new themes were heard.¹⁵

Analysis

Qualitative analysis was conducted using an inductive thematic analysis approach.^16,17 First, at least 2 study team members (S.A.H., L.I.S., I.J.E., A.W.O., or K.M.R.) independently read each transcript and highlighted segments of text (ie, unit of analysis) relevant to the study aims. One team member wrote a preliminary summary of observations for each transcript, and these observations were then reviewed and refined by the larger team in the context of the respective transcript. Consensus observations were coded and analyzed using qualitative research software (NVivo 12, QSR International) and then used, with the interview guide, to generate a preliminary codebook. The codebook was refined as analysis progressed and earlier interviews were recoded as new codes were developed. After all transcripts were coded, 2 team members (S.A.H. and L.I.S.) independently reviewed the coded observations and generated overall themes. Themes were compared by the entire study team and edited and refined using a group consensus process.

RESULTS

Participants (N=20) were, on average, 53.5 years old (standard deviation [SD]: 12.2 years; range: 34–72) and predominantly white (19 of 20, 95%); 1 participant identified as African American. Among the 20 participants, 13 were male (65%) and the remaining 7 were female (35%); 8 participants (40%) had a previous diagnosis of OUD, and 4 of those had a prescription for a MOUD in the prior year. A total of 13 participants (65%) had at least 3 opioid prescriptions in the prior year, with an average of 7.5 prescriptions (SD: 6.6; range: 3–27). Of these, 7 participants had received opioids for acute needs (postsurgery or injuries), whereas 6 were on long-term opioid therapy for chronic pain. One participant had both an OUD diagnosis and 5 opioid prescriptions in the prior year.

Qualitative analysis revealed 6 themes, which are described below with corresponding representative quotes presented in Table 1 (4 patient archetypes), Table 2 (clinic environment), and Table 3 (opioid education).

Table 1.

Example Quotes From Theme 1 (Archetypes of Patient Relationships With Opioid Use and the Health Care System)

Theme	Quote
Archetype 1: Opioids used by patients for chronic pain	“Well, it’s kind of hard to say because I really needed the pills to function. I really did. I was in so much pain.” [58-year-old female] “The abuse, I know, happens. I know there’s a lot of people who abuse it. And that’s really hard for those of us who need it and do everything we can to limit the amount that we take. It’s kind of almost a slap in the face because other people are abusing it.” [58-year-old male] “And like I say, the legitimate users are going to get caught up in the whole push to get this nation off opioids.” [63-year-old male] “It scares me because literally not having that pain medication, or something that would work as effectively, I can’t lift heavy stuff. I can’t make a decent living. So it did scare me because it’s important for me to keep work [sic] and to not feel the pain in my back and shoulders.” [47-year-old male] “I did actually need them. And when I would talk about wanting to go up, [my doctor] said something like, ‘Well, now you sound like an addict.’ And that was really offensive, but at the same time, it was true.” [42-year-old male] “I don’t want to be on any of the drugs … but there’s no alternative. You name it, we’ve tried it.” [64-year-old male]
Archetype 2: Acute use of opioids by patients	“I take them as little as possible. I do not like that feeling. But I’ve had it for a short period of time after brain aneurysm repair; I took it for maybe two days, day and night. And then just two or three nights just so I could sleep after a knee replacement, for maybe a day or two after shoulder repair, and I took it for a day or so after the broken wrist surgery. But I don’t like the feeling when I’m taking it.” [72-year-old female] “As soon as I didn’t need them, I didn’t use them … after the second operation, when they removed the osteo in my leg, I really wasn’t in any more pain … I had already stopped taking them at that point.” [67-year-old male] “Absolutely no worries have I ever had with regard to developing a risk. In fact, I had had a prescription for Tramadol in the past. I had had a prescription for 30 pills, which took me over a year to use. And I kept using this as an example for my responsibility factor in getting on board with another pain medication. And that didn’t seem to have any bearing on anything. And again, I understand there are procedures and policies, et cetera, involved with the entire group of opioid medications. But I felt like I was kind of being treated in a punitive manner in order to be able to just get access to something, in order to be able to assist me with pain that I was having with my back, which was documented.” [61-year-old male] “I didn’t have any worries because I didn’t think I was going to be on them long enough to get addicted to them.” [72-year-old female] “I would probably check out of that conversation [about opioid risks]. Because I never really saw myself that way.” [43-year-old female]
Archetype 3: Problematic opioid use leading to openness to treatment for an opioid use disorder diagnosis	“Well, eventually I needed stronger medications because they weren’t working, and then eventually, my doctor at the time, sent me to a pain clinic. And then they were being monitored, and then we thought it was time for me to go off of them, and I kept getting violently ill every time I’d go off of them. And we just were having a hard time tapering me off of them, and so I just started to do research on my own. And I would ask questions.” [40-year-old female] “I went to my doctor at [clinic name] … to find out how I can get off of opioids, and he referred me to Dr. [name] because Dr. [name] is involved in the program of administering a medication that can help me get off the pills and stop the cravings and the withdraw [sic] symptoms. And so I made my appointment with him. So he prescribed this medication. It just helped me so much immediately. And so yeah, I kind of looked to him as like a savior at this point because it really, really helped me.” [58-year-old female] “I knew I had a problem … I said [to physician], ‘Okay, I want to get off [opioids]. So, if I want to get off them and you have the support system, then I will get off them.’” [42-year-old male]
Archetype 4: Problematic opioid use not yet open to treatment for an opioid use disorder diagnosis	“… they would at least respond with a ‘I’m not ready’ or ‘Can we give it a little longer?’ And there are so many people out there who just want to stay on this stuff [opioids]. And I get that, I do, because it does give some sort of semblance of normalcy.” [39-year-old male] “Now, like I said, I’m an addict/alcoholic, and I seek opioids at certain times in my life because there are certain doctors, word is out … ‘well, go to Dr. so-and-so at such and such a clinic to get your meds.’ I don’t go to [my primary care clinician] for that reason. I go to him for medical reasons only. I’ve never asked him for opioids, and I won’t because there are certain doctors, that’s why you visit them. And then you have your other doctors for medical reasons, which he is.” [66-year-old male]

Table 2.

Example Quotes From Themes 2, 3, and 4 (Clinic Environment)

Theme	Quote
Theme 2: Patients develop their own clinician archetypes	“But a lot of people come up here for the wrong reasons: welfare, free housing, dope doctors. That’s what we call them. And there’s a lot of them still practicing here in the cities like there are, I’m sure, in a lot of other states also.” [66-year-old male] “I think a lot of clinicians just won’t even prescribe [opioids]. And it seems like every doctor or [physician assistant] has their own ceiling about them.” [59-year-old male] “Well, of course [doctors] have different viewpoints. Well, maybe not viewpoints but different methods of working with the patient. I think they each have their own strategies.” [66-year-old female]
Theme 3: Most patients are open to talking about opioid risks but have diverse preferences on how these conversations should be conducted	“Just to talk with the patient in a caring way and basically letting them know that any of this discussion is information. And so it’s not accusatory, it’s just information.” [59-year-old male] “I know some people are a little more delicate when it comes to [talking about opioids], so I would say knowing your patients and being tactful in your approach.” [33-year-old male] “Well, I guess the doctor’s personality, to start with, if he’s just going – if he’s just textbook and doesn’t seem really caring, doing it because he had to, I think that would give me a bad taste.” [66-year-old male] “If the provider was able to approach it with always positive intentions and assuming transparency and honesty, that might make it go better.” [34-year-old male] “I think dependency and addiction to me sound more like risks of the prescription, and misuse sounds more like problems I might have myself, that I might cause or that I might do to myself. So that has a slight accusatory connotation I think to it. Where the other ones feel like side effects that could be unintended. None of them seem bad.” [34-year-old male] “So when you say addicted, it’s such a negative word. And when you say misuse, it seems very deliberate when you’re telling a patient, ‘You’re misusing. You’re deliberately doing something wrong.’” [60-year-old female] “I used to feel really badly about that word [addict], actually, because I had always used it for my mom and people who I thought put themselves in a situation to become an addict. So I actually prefer the word dependent -- opiate dependent over addict, for me – to refer to me. … I guess I would rather say, ‘You can become a dependent.’ Those wordings, I guess, I like better.” [40-year-old female] “Addiction would give me a bad feeling.” [59-year-old male] “[Conversations go poorly] if there’s more of a blaming attitude to it or an assumption. If that were me, I feel distance between the doctor and myself, threatened.” [66-year-old female] “I guess for the relationship that I have with [my PCC], I go see him about everything, so if he wanted to talk about opioids he could.” [42-year-old male] “I understand [opioid problems] … coming from a family of people with addiction issues, it wouldn’t faze me personally. I would understand what they were trying to convey.” [47-year-old male] “Just by the way they would speak about [opioids] and it made me feel like if I had to take them that I’m some sort of, I don’t know, lower than them or just that, some sort of criminal or something.” [59-year-old male]
Theme 4: Primary care is an appropriate setting for these discussions	“Oh, my primary care. I trust him and his judgment more than anything.” [58-year-old male] “I think it’s 100% important [for doctor to have these conversations]. And I think all of your doctors should be in contact with each other, especially your primary care doctor should know who your other doctors are in all of your care, and any medications they’re prescribing you, like a psychiatrist or somebody doing methadone maintenance. Any doctor that’s involved in your health care, your primary should know all of them and what they’re prescribing.” [40-year-old female] “I would say it’s extremely important [for primary care doctors to talk about opioid risks]. And then your primary care physician is your interface with the system. So this is the person you rely upon for medical advice. And they know you. They have a relationship with you.” [67-year-old male] “First of all, I trust [my PCC]. I trust her ability to advise me in any fashion. Plus, she has an overall view of what I’m going through with respect to any degree of my medical history.” [61-year-old female] “And they should say, ‘Hey, you need to go see a specialist – a pain doctor – and let them do what they do to figure out how much your dosage should be.’ I don’t think a primary care one should be – he’s only got limits. That’s why they have specialists.” [64-year-old male]

Table 3.

Example Quotes From Themes 5 and 6 (Opioid Education)

Theme	Quote
Theme 5: Patients have limited awareness of opioid rescue medications and medications for opioid use disorder	“No. Nobody’s mentioned [Narcan].” [66-year-old female] “I’ve never heard of [Narcan].” [43-year-old female] “[Narcan] doesn’t ring a bell, no.” [34-year-old male] “Another thing I think that would help addicts is a lot of them don’t know about the medications that help you stay off opioids … If I would’ve known that there were medications out there, I possibly would’ve talked to my doctor about it.” [58-year-old female]
Theme 6: Handouts are acceptable when they come from the clinician	“Fine … I think [the messages on the handouts are] pretty routine. And so I didn’t have any emotional, positive or negative, just seemed kind of routine.” [34-year-old male] “Oh [getting a handout would] be fine. Like I was saying, [my doctor]’s honest, I value his opinion, and he elicits the conversation out of you to answer questions that he needs answered, and I need answered, too.” [63-year-old male] “Well, personally in this moment in this body in this soul, I would be fine with [getting a handout]. I’d be like, ‘Oh, that’s okay, I’ll read that.’ But I’m pretty easygoing. There are a lot of people out there who’d probably be like, ‘What the hell? What are they trying to say?’ There could be people who take it wrong. There’s [sic] so many not-balanced people out there that it would be offensive coming from a computer, not a human.” [39-year-old male] “Well, the computer doesn’t know me. The computer’s working off of information that is input into it by someone who doesn’t know me.” [33-year-old male]

Theme 1 — Patient Relationships With Opioid Use and the Health Care System Can Be Understood Using an Archetypal Heuristic (Table 1)

Archetypes do not reflect a definitive type of a patient or relationship, but characteristics that were commonly found together. Thus, archetypes can provide an initial indicator PCCs may consider when engaging with their patients about opioids and OUD risk. Patients approached the topic of opioids differently depending on their experience and relationship with opioids and the health care system. Our sample yielded 4 archetypal groupings characterizing patient relationships with opioids (Table 4) that varied in perceptions of their perceived risks for problems with opioids, their indication for using opioids, their views about why others might use opioids, their beliefs about whether they have a substance use disorder, and their openness to treatment for OUD.

Table 4.

Characteristics of Patient Archetypes

Archetype	History of long-term opioid use	Open to substance use disorder treatment	Self-perceived risk of problems with opioids	Use of opioids perceived as necessary to manage pain
1. People with chronic pain	Yes	No	No	Yes
2. People who use opioids acutely	No	No	No	Yes
3. People with opioid use disorder and open to treatment	Yes	Yes	Yes	No
4. People with opioid use disorder and are not yet open to treatment	Yes	No	No	No

Archetype 1: A self-recognized dependence on opioids perceived as necessary for functioning in daily life, often to manage chronic pain

This archetype fits well with patients who reported that they have used opioids for a long term and have not found viable alternatives to manage their pain. Patients with statements consistent with this archetype acknowledged a dependency on opioids to function. However, they also characterized themselves as different from people with OUD and worried system-level policies to reduce opioid prescribing would negatively affect them. They interpret actions taken by health systems that treat them as though they have a substance use disorder as offensive. Furthermore, they often blame the difficulties in obtaining opioid treatment on people with OUD.

Archetype 2: An acute use of opioids, often after surgeries or procedures, with a corresponding self-perception of low risk for harm or addiction

Patient responses consistent with this archetype described short-term, acute opioid use that stopped when they were no longer necessary to manage pain. They do not see themselves as at risk for problems with opioids and believe there is limited addiction potential with their short-term use. Despite their beliefs that they have low likelihood of developing addiction, some patients noted frustration with being able to get opioids when they feel they are needed. Because they see themselves as low risk, some patients may not fully engage in that conversation.

Archetype 3: Opioid use patterns self-recognized as problematic and leading to serious impairment or distress and accompanied by an openness to treatment

Patient statements consistent with this archetype acknowledged their diagnosis of OUD and perceived treatment as necessary.

Archetype 4: Opioid use consistent with a pattern that leads to serious impairment or distress that, whether self-recognized or not, is not accompanied by an openness to treatment

Patient statements consistent with this archetype expressed a desire to hide their problematic opioid use and seek out different PCCs who are known to be more liberal in prescribing opioids. They use opioids to satisfy a craving or a use disorder.

All patients who made statements consistent with the chronic pain on long-term opioids archetype and most patients (5 of 6) who made statements consistent with the acute prescription archetype were identified for the study based on having ≥3 opioid prescriptions in the past year. All patients with OUD on a MOUD made statements consistent with the OUD and open-to-treatment archetype. Of 4 people with an OUD diagnosis and no MOUD, 2 made statements consistent with the OUD-not-yet-open-to-treatment archetype. The only patient with ≥3 prescriptions and an OUD diagnosis did not clearly fit with the acute prescription archetype (Archetype 2) or the OUD-not-yet-open-to-treatment archetype (Archetype 4). Similarly, 1 patient with an OUD diagnosis and no prescriptions (opioids or MOUDs) made statements consistent with both the acute prescription (Archetype 2) as well as the OUD-not-yet-open-to-treatment (Archetype 4) archetypes.

Theme 2 — Patients May Develop Archetypes of PCCs, Characterized by Perceptions of Each PCC’s Role in Their Care (Table 2)

Patients described developing a sense of a PCC’s willingness or unwillingness to prescribe opioids. As noted in the description of Archetype 4 (problematic opioid use that is not treated), some patients will only see their PCCs for medical reasons but will seek “dope doctors” or PCCs who more freely prescribe opioids.

Theme 3 — Patients Are Open to Talking About OUD Risk but Have Diverse Preferences for (A) How Conversations Should Be Conducted, (B) Who With, and (C) What Terminology Is Used (Table 2)

Analysis of our sample suggested that patients preferred talking about opioid risks with a trusted PCC or one who is about to prescribe opioids for them. These conversations should be “gentle” (ie, come from a place of empathy and compassion without judgment), emphasize information, and respect patient autonomy. Patients stated PCCs should approach these conversations with a caring attitude, transparency, and honesty. Many patients said conversations go poorly when PCCs talk down to them or use accusatory tones. Ultimately, patients recommended PCCs emphasize information-sharing and be nonaccusatory when talking about opioid risks.

Patients varied in their comfort with different terminology. Some patients noted feeling uncomfortable with words like “addiction” and “dependence.” Other patients did not share these negative feelings and emphasized how their personal or professional backgrounds shaped their views. Patient reactions to specific words may correlate with their best-fitting archetype as well as their personal and professional experiences with opioids and substance use disorders.

Theme 4 — Primary Care Is an Appropriate Place to Discuss Opioid Risk Because PCCs Are Trusted Experts Who Know Their Patients Well (Table 2)

Many patients stated they preferred their PCCs discuss opioids with them because of their trusting relationships and some expected PCCs to discuss opioid risks regardless of whether opioid medications were to be prescribed. Further, patients thought PCCs should discuss opioid risk because they manage a person’s overall health. Two patients noted the limits of PCCs’ abilities, identifying specialists (eg, pain management) as more appropriate for these discussions.

Theme 5 — Patients Have Limited Awareness of Opioid Rescue Medications and MOUDs (Table 3)

In response to questions about whether PCCs had talked to them about having rescue medications at home, half of the patients said they had never heard of them before. Despite limited awareness of rescue drugs like naloxone or MOUDs like buprenorphine, patients at increased risk for problems with opioids do think PCCs should be up front about offering these options. Another patient noted she was unaware that MOUDs were available prior to starting them and thinks PCCs should raise awareness of the availability of these medications.

Theme 6 — Handouts May Be More Welcomed by Patients if Perceived to Come From PCC Judgment Rather Than a Computer Algorithm (Table 3)

Patients were generally accepting of receiving a handout encouraging them to discuss opioid risks with their PCC during their visit. Several patients expressed general acceptance of receiving a handout from a PCC; however, when asked if they knew a computer algorithm had generated the handout for them, they were more suspect because the computer does not “know” them like a person does.

DISCUSSION

This study aimed to learn how patients view conversations with a clinician about opioid risks in primary care to help PCCs think about their approach to these discussions in a more organized and sensitive way. Analysis of the sample identified 6 themes that could help in addressing patients that PCCs may see; the appropriateness of the setting, conversation, and terminology; and opioid education needs. These themes extend the literature on opioid risk conversations by including diverse perspectives from patients who may be at risk for problems with opioids, including people with chronic pain on long-term opioid therapy, people who use opioids acutely, and people with OUD who are and are not receiving treatment. Most patients felt conversations with PCCs were acceptable but preferred those conversations came from a place of empathy and compassion and avoided stigmatizing or judging patients.

Our findings suggest that a heuristic with at least 4 archetypes may assist PCCs in having effective conversations with patients about opioid use risks, particularly for persons with chronic pain on long-term opioid therapy who are at greater risk of OUD and harm from opioids.¹⁸ Consistent with previous research, individuals with characteristics consistent with this archetype may object to being labeled as someone with a substance use disorder.¹³ Further, they report frequent experiences of being stigmatized by PCCs and having their experiences invalidated.^19,20 PCCs working with patients making statements consistent with this archetype may consider how their actions to mitigate opioid risks may be perceived instead as threatening. Emphasizing personal concern for patient safety and validating patients’ experiences of pain may be particularly helpful with patients in this archetype.

Across archetypes, patients wanted to be treated with respect and compassion and to feel heard. In general, patients recommended that PCCs approach conversations about opioid risks with genuine concern, taking patients’ specific situations into account. Patients want their PCCs to listen to their perspectives before switching to a new treatment plan.²¹ Patient-clinician interactions are deemed to be more collaborative when PCCs are more compassionate, patient-centered, nonjudgmental, and validating of patients’ concerns.⁹

PCCs should be reassured that patients with elevated risk for opioid problems feel that primary care is an appropriate setting in which to discuss opioid risks, especially with trusted clinicians. Further, patients find it acceptable to receive a handout prompting a discussion about opioid risks from their PCC. Patients who trust their PCCs are more accepting of limitations on opioids²² and feel less stigmatized.¹³ Patients are more likely to trust PCCs who demonstrate care, empathy, and respect.²³ Thus, patients trust PCCs who display a level of interpersonal competence and are more likely to follow their clinical recommendations. However, repeated negotiations about opioid use can erode those trusting relationships.²⁴ Continued listening and collaboration are key to maintaining positive patient-PCC relationships.

Our findings indicated that personal comfort with different terms varies across patients and archetypes. Most people objected to stigmatized labels (like addict or addiction) when the context was in reference to themselves. Generally, patients did not think that the term “dependence” carried the same stigmatizing weight, indicated that the behavior may not be entirely controllable, and was universally understood. As clinical terms become more widely used among the lay population (eg, OUD), it will be important to learn more about how patients respond to those terms.

A potentially novel finding was that some patients have their own archetypes about PCCs’ comfort in prescribing opioids. These patients learn which clinicians in the community are more liberal in their opioid prescribing and see them primarily to obtain opioids while avoiding the topic with their own PCC. Other patients also described their personal PCCs as being overly cautious about prescribing opioids or being willing to work with them based on their medical history. The PCC schemas that patients develop may ultimately affect their interactions with PCCs, including what information they are willing to share with them.

Study participants generally felt like they were at low risk for problems with opioids and were not aware that overdose rescue medications or MOUDs were available options. This is consistent with other research that demonstrates that most people feel like they are low risk for problems with opioids, such as an overdose, but believe clinicians should offer naloxone.^25,26 Thus, PCCs should be aware that some patients may not personalize opioid risks when having these conversations but that they should still inform patients of the availability of MOUDs and naloxone.

Limitations

This study was strengthened by a robust qualitative analytic procedure and a sample that represented different people who may be targeted for opioid risk discussions. Its heterogenous target population also contributed to variability in responses and allowed us to describe the archetypes that we found. However, this study was limited by recruitment of patients using primary care in a single health system, which may limit the transferability of responses to patients in specialty settings or from other health systems or geographic locations. Further, this is a small, nonrandom sample of people receiving care from 1 of 8 PCCs in an integrated health system who agreed to be interviewed; although we reached saturation in our qualitative themes within these patients, there may be perspectives of people who were not included in this sample. Finally, this sample was predominantly white, not Hispanic, and middle-older age, which limits the generalizability of these experiences from patients who identify as members of other racial or ethnic groups or who are younger.

CONCLUSIONS

It is critical for primary care clinicians to address opioid risks with their patients, yet conversations about opioid risks and potential opioid use problems can be challenging for both patients and PCCs. Our study suggests that patients feel these conversations are improved when PCCs approach them with compassion and consider what they know about the patient to tailor the conversation to patient needs. Future research is needed to assess the applicability of these themes in a larger sample of patients and to observe actual interactions between clinicians and patients about this topic.

Patient-Friendly Recap

• The U.S. opioid epidemic is pervasive, yet both patients and primary care clinicians struggle to discuss opioid risk. Study authors interviewed 20 “high-risk” patients to get their perspectives on having conversations about opioid screening and treatment at clinic visits.

• Patients’ use of opioids fell into 4 distinct types, each requiring a different conversational approach. Still, most patients believed that primary care is an appropriate setting for compassionate opioid risk discussions, which should be tailored to a patient’s specific needs.

• Patients have limited awareness of available overdose and opioid use disorder medications.

• Larger-sample and real-world research into these themes is needed to assess applicability in actual patient-clinician interactions.