Abstract
Objective
To describe patient-provider communication about opioid pain medicine and explore how these discussions affect provider attitudes towards patients.
Methods
We audio-recorded 45 HIV providers and 423 patients in routine outpatient encounters at four sites across the country. Providers completed post-visit questionnaires assessing their attitudes towards patients. We identified discussions about opioid pain management and analyzed them qualitatively. We used logistic regression to assess the association between opioid discussion and providers’ attitudes towards patients.
Results
48 encounters (11% of the total sample) contained substantive discussion of opioid-related pain management. Most conversations were initiated by patients (n=28, 58%) and ended by the providers (n=36, 75%). Twelve encounters (25%) contained dialogue suggesting a difference of opinion or conflict. Providers more often agreed than disagreed to give the prescription (50% vs. 23%), sometimes reluctantly; in 27% (n=13) of encounters, no decision was made. Fewer than half of providers (n=20, 42%) acknowledged the patient’s experience of pain. Providers had a lower odds of positive regard for the patient (adjusted OR=0.51, 95%CI: 0.27–0.95) when opioids were discussed.
Conclusions
Pain management discussions are common in routine outpatient HIV encounters and providers may regard patients less favorably if opioids are discussed during visits. The sometimes-adversarial nature of these discussions may negatively affect provider attitudes towards patients.
Practice Implications
Empathy and pain acknowledgement are tools that clinicians can use to facilitate productive discussions of pain management.
Keywords: communication, pain, mixed methods
INTRODUCTION
Chronic pain affects approximately 76.2 million Americans--more than diabetes, heart disease and cancer combined1. Some of these patients seek the expertise of health care professionals to relieve their pain, often through pharmacologic means. In recent years opioid pain medications have become more frequently used to treat patients with chronic and severe pain2. Consequently, pain management is now a frequent topic of conversation between patients and their healthcare providers, with a recent study from the National Center for Health Statistics finding that analgesics were the most common type of therapeutic drug mentioned in outpatient visits3.
It has also been well documented that HIV positive patients have a high prevalence of chronic pain associated with significant functional impairment4. Furthermore, there is evidence to suggest that pain is severely undertreated in HIV-positive patients, especially among those with a history of intravenous drug use5. Chronic untreated pain can lead to depression, psychological distress, and impaired quality of life, particularly in HIV-positive patients6.
Considering the balance between treating pain to improve quality of life and avoiding the harmful consequences of medications, it is essential that patients and providers communicate effectively about opioid pain management. However, evidence suggests that these discussions are sometimes a source of frustration and conflict. A recent study showed that 73% of providers described working with chronic pain patients to be a “major source of frustration”7. One physician described working with this population as a “thankless task”, asking “who wants to be confronted with failure every day?”8. Providers may find it difficult to communicate effectively with these patients for at least three key reasons. First, high rates of opioid diversion and misuse9 lead providers to be skeptical of patients requesting opioids. Second, providers may feel disillusioned working with patients whose chronic pain they are unable to adequately treat. Finally, discussions about opioids can develop into battles for control that turn providers and patients into adversaries rather than collaborative partners10. Without a collaborative relationship with their provider, patients often struggle to maintain their credibility and leave visits feeling “rejected, ignored, and belittled, blamed for their condition”11.
Although some studies have surveyed patients and providers about their general experiences regarding opioid pain management, few studies have examined patient-provider communication during actual encounters. One recent study characterized communication about opioids with chronic pain patients in primary care visits12. However, there is little known about the content of these discussions with HIV-positive patients in particular, or about how having these discussions may affect provider attitudes. The purpose of our study is two-fold. First, we explored how opioid discussions affect provider attitudes towards patients. We hypothesized that providers would have more negative attitudes towards patients whose visits contained discussions about opioid pain medications. Second, we sought to quantitatively and qualitatively describe the content of patient-provider communication about opioid pain management.
METHODS
Study Design, Subjects, and Setting
We conducted a mixed methods secondary analysis of observational data collected as part of the Enhancing Communication and HIV Outcomes (ECHO) Study13,14,15,16,17,18. Study subjects in this analysis were 45 providers and 423 of their HIV-infected patients at four primary HIV care sites across the United States (Baltimore, Detroit, New York, and Portland, OR). The study received IRB approval from each site. Eligible providers were physicians, nurse practitioners, or physician assistants who provided primary care to HIV-infected patients. Overall, 82% of all providers across the four sites participated. Eligible patients were HIV-infected, older than 18, English-speaking, and had had at least one prior visit with their provider. Overall, 73% of approached eligible patients participated. The most common reasons for patient refusal were that they did not have time to complete the interview (65%), that they were not feeling well (13%), and that they were not interested in studies (8%).
Data Collection Methods
HIV providers who agreed to participate completed a baseline questionnaire. On the baseline questionnaires, providers supplied basic demographic information (age, sex, type of training). Research assistants approached patients of participating providers in the waiting rooms, with the goal of enrolling ten patients per provider. Only patients presenting for follow-up visits were included in the study. After patients gave informed consent, research assistants collected digital audio-recordings of clinical encounters and conducted a post-visit interview with patients. In post-visit interviews, patients’ self-reported demographic information (age, sex, race/ethnicity, employment, and education), depressive symptoms19, and illicit drug use20.
After each encounter, providers completed post-visit questionnaires containing measures regarding their attitudes towards the patient they had just seen, including frustration (using a single item developed for this study, “this patient frustrates me”), affiliation (using a single previously-validated item, “I could see myself being friends with” this patient”)21, and provider regard for the patient.22,23 Provider regard was determined by combining the following five items, each measured on a five-point Likert scale (‘strongly agree’ to ‘strongly disagree’), and each beginning with “Compared to other patients,”: 1) I have a great deal of respect for this patient’, 2) ‘I really like this patient’, 3) ‘I find this patient very interesting’, 4) ‘, I find it easy to understand this patient’, 5) ‘this patient is one of those people who makes me feel glad I went into medicine.‘
Identification of Dialogue about Opioid Pain Medication
Audio recordings were transcribed verbatim. Transcripts containing discussions of opioid pain medicine were identified by searching the documents for a list of 20 opioid-related terms that was developed with the input of investigators from each study site. Search terms included both the names of specific opioids (e.g. Oxycontin, Percocet, etc.) as well as more general terms (i.e. “pain med”) in order to improve the sensitivity of our search. Transcripts containing at least one search term were reviewed, and were excluded from analysis if they did not contain any discussion of opioid pain management (e.g. the term ‘oxy’ could refer to ‘oxygen’) or the discussion was limited to a simple medication reconciliation (e.g. “Are you still taking the Percocet?”, “Yes”). If the transcript contained any discussion of pain or pain medication beyond this type of medication reconciliation, we considered it ‘substantive’ discussion.
Quantitative Analysis of Encounters with Discussion of Opioid Analgesics
We used descriptive statistics to assess the characteristics of the entire study sample. We created a binary variable representing whether or not the encounter contained a discussion of opioids based on our method of dialogue identification. We used chi-squared and Wilcoxon-Mann-Whitney tests to compare encounters with and without discussions of opioids, as well as with and without the presence of conflict. We compared encounters with regard to patient characteristics (self-reported age, sex, race/ethnicity, employment status, education, length of relationship with provider, depressive symptoms, and illicit drug use), and provider characteristics (age, sex, and type of training).
Due to skewed response distributions, we dichotomized our three outcome variables (provider frustration, affiliation, and regarding). Given that providers more frequently reported positive attitudes than negative attitudes towards their patients, we grouped neutral responses with those that indicated negative attitudes. Thus, provider frustration with patient was dichotomized so that neutral responses were grouped with ‘agree’ or ‘strongly agree’. Provider affiliation with patient was dichotomized so that neutral responses were grouped ‘disagree’ or ‘strongly disagree’. Provider regard scores were grouped into tertiles and a dichotomous variable was created to compare high regard versus middle or low regard.
Finally, we used multiple logistic regression for multivariate analysis of provider attitudes comparing encounters with and without pain management discussions. All logistic models adjusted for study site and accounted for clustering of patients within providers using generalized estimating equations24. We considered for inclusion in multivariate models any patient or provider variable that was associated with both the presence of pain management discussion and any of our outcome variables, to a statistical significance of p < 0.2 (patient gender and employment status). We also included two variables thought to be potential confounders based on the investigators’ clinical experience (patient depression and active illicit drug use). We generated an interaction term to test whether illicit drug use modified the effect of the discussion of opioids on provider attitudes towards patients. We performed additional logistic regression analyses to determine if provider attitudes differed when opioid medications were mentioned briefly during medication reconciliation but not discussed. All analyses were conducted using Stata Version 1125.
Coding of Encounters with Discussion of Opioid Analgesics
After identifying transcripts containing discussions of opioid pain medicine, we read through each encounter and identified relevant dialogue including important contextual elements. Two investigators (HK, MCB) determined thematic categories about pain medications within the transcripts, primarily focusing on the content and structure of the discussion. We met to discuss and modify these categories, eventually reaching a final list of important contextual factors and overarching themes. We had no theoretical framework at the start of this analysis. Through a process of open inductive coding, we identified and coded the following aspects of the dialogue: who initiated the discussion (patient or provider), what the nature of the pain was, whether there was a diagnostic work-up related to the underlying source of pain, whether the provider made an empathic statement acknowledging the pain, whether the provider expressed concern about the use of opioids by the patient, whether the provider suggested an alternative treatment, whether there was disagreement, conflict, or anger within the discussion (confirmed by listening to audio-recorded visits), whether the provider prescribed opioid medication, whether the provider shifted the topic away from pain medication use, and whether the provider shifted decision making to other clinicians. One coder (HKH) applied this scheme to all relevant transcripts, and a second coder (MCB) reviewed all coding. Disagreements were discussed until consensus was achieved. We describe the frequency of appearance of these conversational elements, and provide examples below. Finally, because dialogue containing conflict seemed the most difficult, we relate several of these conversational elements together to briefly describe the nature and outcome of these discussions.
RESULTS
Study Sample and Prevalence of Discussions of Opioid Analgesics
Patient and provider characteristics of the entire sample are presented in the first column of Table 1. Our patient sample was predominantly male, predominantly non-white (58% African-American, 14% Latino, 5% other), and had a low level of employment (26%). A significant proportion of patients classified themselves as currently using illicit drugs at the time they were surveyed (28%). Providers were predominantly female, white, and trained as physicians as opposed to nurses or physician assistants.
Table 1.
Patient and Physician Characteristics by Whether Opioids Were Discussed During Encounter
| Entire Sample | Opioids Discussed |
Opioids Not Discussed |
p-value | |
|---|---|---|---|---|
| n=48 | n=375 | |||
| Patient Characteristics | ||||
| Age, Mean | 45.43 | 45.98 | 45.29 | 0.856 |
| Sex, % Female | 35% | 45% | 34% | 0.140 |
| Race, % White | 23% | 30% | 22% | 0.230 |
| Education, % High School Graduate | 72% | 64% | 73% | 0.187 |
| Depressive Symptoms, % Highest Tertile | 33% | 36% | 32% | 0.659 |
| Employment, % Working | 26% | 17% | 27% | 0.148 |
| Length of Relationship with Provider, % > 5 yrs | 66% | 60% | 67% | 0.315 |
| Illicit Drug use, % Active | 28% | 34% | 27% | 0.310 |
| Provider Characteristics | ||||
| Age, Mean | 44.60 | 45.53 | 44.39% | 0.297 |
| Sex, % Female | 57% | 57% | 57% | 0.948 |
| Race, % White | 69% | 77% | 68% | 0.313 |
| Training, % Physician | 73% | 70% | 73% | 0.747 |
Figure 1 describes the identification of encounters containing discussions of opioid pain management. Approximately one-third of transcripts contained at least one opioid-related search term. Forty-one of these transcripts were eliminated from analysis because the use of the search term in that encounter did not pertain to treating pain with opioids. For example, the search term ‘methadone’ yielded 19 transcripts in which methadone maintenance therapy for drug addiction was discussed, but not the use of methadone for treating the patient’s pain. Another 41 transcripts were eliminated from analysis because the search term was used in the context of reconciling medication lists during the encounter and not in a substantive discussion between patient and provider (e.g., “D: How much of the Oxycontin are you taking? P: Two 40 milligrams twice a day”). After excluding these encounters, we found that substantive discussion of opioid pain medicine occurred in 11% of encounters (48/423) with 62% of providers enrolled in the study (28/45).
Figure 1.
Selection of Transcripts Containing Substantive Discussions of Opioid Analgesics
Association of Opioid Discussion with Patient and Provider Characteristics
In unadjusted analysis, there were no significant associations between patient or provider characteristics and the presence of a discussion about opioid pain management in the encounter (Table 1). Table 2 shows the frequency of provider attitudes, and the association of provider attitudes with the discussion of opioid pain medicine. In post-encounter surveys, 30% of provider responses in the entire study sample indicated frustration with the patient they had seen during the visit, 32% indicated that they could see themselves being friends with the patient, and (by definition) approximately one-third (36%) of provider responses rated the patient in the top tertile of regard.
Table 2.
Association of Opioid Discussion with Provider Attitudes towards Patient
| Provider Attitude | Unadjusteda OR |
(95% CI) | Adjusteda,b OR | (95% CI) |
|---|---|---|---|---|
| Frustrationc | 2.13 | (1.03–4.43) | 1.84 | (0.87–3.91) |
| Affiliationd | 0.74 | (0.36–1.50) | 0.76 | (0.38–1.53) |
| Positive Regarde | 0.50 | (0.27–0.93) | 0.51 | (0.27–0.95) |
Comparing encounters with discussion of opioids to those without discussion of opioids, all analyses (adjusted and unadjusted) account for study site and clustering of patients within providers using GEE.
Additionally adjusted for patient gender, employment, depressive symptoms, and illicit drug use.
Odds ratio referenced against provider strongly disagrees or disagrees with the statement “this patient frustrates me”.
Odds ratio referenced against provider strongly disagrees or disagrees with the statement “I could see myself being friends with this patient.
Odds ratio referenced against low provider regard.
When opioids were discussed versus not discussed in the encounter, there was a greater than two-fold higher unadjusted odds of providers’ indicating frustration with patients (OR: 2.13, 95% CI: 1.03–4.43, p=0.04), although this association was not significant after adjusting for patient and provider demographic factors. When opioids were discussed in the encounter, there were lower odds that providers would rate the patient in the top tertile of positive regard (OR: 0.50, 95% CI: 0.27–0.93, p: 0.03). Opioid discussions remained independently associated with lower regard after adjusting for patient factors including patients’ active illicit drug use. The presence of a conversation about opioid pain management was not associated with the provider’s reported affiliation with the patient. When pain medications were mentioned briefly in the context of medication reconciliation but not discussed (n=41), there were no significant differences in provider attitudes towards patients.
Characteristics of Discussions about Opioid Pain Medicine
Patients initiated most conversations about opioids (58%). The most common types of pain discussed were musculoskeletal/back (31%) and neuropathic (17%); in 29% of encounters, the type of pain was never mentioned. In 43% of encounters, the underlying etiology of pain was obvious from the discussion (with further diagnostic evaluation probably unnecessary in the opinion of the investigators) or there was dialogue indicating a diagnostic workup (e.g. provider questioning about the nature of the pain); the majority (57%) did not include any diagnostic evaluation.
Additional characteristics of these discussions are summarized in Table 3. Fewer than half of the discussions (n=20, 42%) included the provider stating an explicit acknowledgement of the patient’s pain. Provider’s acknowledgement of pain did not always result in patients’ receipt of an opioid prescription. Of those 20 encounters in which providers made an explicit statement acknowledging patients’ pain, opioids were prescribed in approximately half (n=11). In almost half of the encounters, providers verbalized concern about the opioid regimen (46%), often by suggesting dosing changes or describing risks and side effects of potent opioids. Providers often suggested alternative therapeutic options (48%) such as non-steroidal anti-inflammatory drugs (NSAIDS) or heating pads. In 5 visits (10%), patients acknowledged that they were receiving pain medication illegally, which was not addressed by the provider in 3 out of 5 instances.
Table 3.
Characteristics of 48 Discussions about Opioid Analgesics
| Category | n (%) | Examples | |
|---|---|---|---|
| Empathy | |||
| Provider verbally acknowledged pain | 20 (42%) | D: I mean, I’m sorry that all these things hurt. | |
| P: mhm | |||
| D: um And we’ll try to work on pain medicine and maybe, eventually other things- | |||
| P: mhm | |||
| D: to get better, better improvement like if we can get your knee really treated properly then maybe, I know- | |||
| P: I’ll dance again. | |||
| _______ | |||
| P: I still need me some. Can I get a few? | |||
| D: What do you use the Vicoden for? | |||
| P, For my legs and stuff | |||
| D: Oh, for your arthritis and your knees….You’ve been evaluated and you have bad osteoarthritis in the knee – | |||
| P: Yeah, that’s why | |||
| D: And that is painful and I know you can’t take Motrin so. | |||
| P: No, that pill is bad for my kidneys. | |||
| D: Right, so the Vicoden is fine. | |||
| P: Yeah, they help me. | |||
| D: Alright | |||
| Provider did not verbally acknowledge pain (‘missed opportunity for empathy’) | 28 (58%) | P: I mean it wakes me up in the middle of the night cryin’ it hurts so bad. | |
| PC: she walk in I help her put her underwear on and stuff, she can’t take a shower it hurts so bad sometimes and stuff. | |||
| D: Well, um yeah, I’m gonna send you to the bone doctors and and we’ll be able to get the Oxycontin based on this new diagnosis, now that we have the MRI resolved. | |||
| P: OK | |||
| _______ | |||
| P: The pain in the shoe is killing me. It won’t leave me alone. | |||
| D: Say Ahh. | |||
| P: Ahh. | |||
| D: Lift up your tongue…When was your last dentist appointment? | |||
| Concern about Opioid Use | |||
| Provider verbalized concern about opioids or suggests reduction in use | 22 (46%) | D: And now I also want you to think about coming, backing off that a little bit if you can. | |
| P: Yeah well I can't do everything all at once. | |||
| D: No I know but it's just a long-term goal, okay? | |||
| P: That's a long-term goal. | |||
| ________ | |||
| P:“I can’t just stop takin’ ‘em.” | |||
| D: “Why?” | |||
| P: “What do you want me to do, go crazy?” | |||
| D: “You could slow down. Say go from four to three.” | |||
| ________ | |||
| D: My feeling is- | |||
| P: you know how I feel. | |||
| D: you know how I feel | |||
| P: Yeah, I know how you feel. | |||
| D: And you know why I feel that way. | |||
| P: That’s how I’m gonna get drugs, okay. | |||
| D: And you know it’s not that I don’t trust you, it’s your illness that I’m concerned about and we know that when people start to use again, that they take poor, worse care, I should say, of their HIV and I don’t want that. | |||
| Alternative Treatments | |||
| % Provider suggested an alternative therapy | 23 (48%) | D: I will give you some ibuprofen … if that’s not controlling it then you need to let me know, we can switch around to another and you can try something else and see if that works better … | |
| ______ | |||
| D: So, well you’re gonna see a neurologist…We can start you on something. What you are describing sounds somewhat like neuropathy. We can start you on some Neurontin if you want. | |||
| P: What’s that? | |||
| D: That’s something that works on neuropathy, it works on the nerves actually. It’s a medication. | |||
| P: What about my shoulder? | |||
| D: You can have some Naprosyn for it and some patches, like lidocaine on top if you want. | |||
| P: What’s that? | |||
| D: That’s just to numb it. You know like they give you something to numb your tooth and that’s the same thing. You can try that. Let’s start with that. Let’s try that and then I would want to get an MRI of your shoulder. | |||
| Conflict | |||
| Conflict or difference of opinion regarding pain treatment | 12 (25%) | P: You know, I don't even want to come here anymore. I wanna – I wanna – I wanna change it. I wanna – I'm tired of this thing‥…Now all of a sudden I'm not on medications? I been with you for the – for ten years -- | |
| D: [Name of patient], you haven't been on meds – I haven’t given you any meds. | |||
| P: Ma'am, no, that's a lie, that's not true Miss [name of doctor]. I'm not an idiot and I'm and then -- | |||
| D: Well I don't know how you got them -- | |||
| P: And then you people keep treating me like an idiot here. | |||
| D: Because – look, look, look -- | |||
| P: Now they keep telling me, "Oh, you don't take Oxycontin." "What do you mean?" I been Oxycontin, I'm on Oxycontin for two years because you didn't give it to me for two prescriptions of fifteen days, all of a sudden now I never take that medication. The other medication is no good, it doesn't work. My feet are still inflamed, I'm sick of feeling like that. You know, come on, this is ridiculous. I need to get to a place where I need to get my stuff taken care of. I can't keep like this because everybody wants to change me, or cuz you feel I'm doin' drugs or whatever you don't want to give me medications. I don't want to hear that. It's not fair to me. | |||
| D: Okay. There are two different issues here, [name of patient]. | |||
| P: No there's not. There's no different issues. | |||
| Prescribing Opiods | |||
| Prescribe opioid, sometimes reluctantly or for a limited time | 24 (50%) | D: I’ll give you a little bit of oxycodone but you need to get off that, okay? | |
| ________ | |||
| D: One more month of Vicodins, all right? And then, …you know I don't want you to be in pain, right? I am just very wary of you getting back up on your Vicodins again. | |||
| Changing Topic | |||
| End discussion by changing topic, sometimes abruptly | 36 (75%) | P:“The end of the night, where I sleep, my back hurt me real bad…that’s when I take the Percocet and…” | |
| D:“Let me ask a quick question for ya” | |||
| P:“Uh Huh” | |||
| D:“When was the last time you had a flu shot? | |||
| _________ | |||
| D: I want you to consider going to the pain thing. | |||
| P: Uh uh. I just want my regular prescription what you all been giving me every month. And I -- | |||
| D: Your last hepatitis C viral load was 1,560,000. | |||
| Shifting Decision Making | |||
| Defer decision making to other care providers | 5 (10%) | D: Yeah – we don't give – I told you they don't pay for Oxycontin and we don't give Oxycontin anymore. | |
| P: Ma'am I'll pay for it myself. I don't need you guys to pay for it. | |||
| D: No, I can't – I – you know, that goes through Dr. [name of doctor] to – the neurologist. | |||
| P: No, other people take it – I don't understand what the problem is. | |||
| __________ | |||
| P: When they gonna be able to, I can’t get me no Vicoden. | |||
| D: What do you use – | |||
| P: I used to get ‘em until what happened. | |||
| D: Well we deferred the prescription of pain medications to the primary doctors – | |||
| P: Them guys full of shit | |||
| D: And we’ll concentrate on the HIV. | |||
P=Patient; D=Doctor: PC=Patient’s Companion
Twenty-five percent of encounters (n=12) contained dialogue suggesting difference of opinion or conflict. Sometimes these discussions contained evidence of patient anger or provider frustration (see conflict example in Table 3). The outcome of encounters containing a disagreement or explicit conflict between patient and provider are shown in Figure 2. In the majority of the 12 discussions containing conflict, the physician ultimately agreed to prescribe an opioid (n=7), often after a lengthy discussion (as indicated by more than fifty turns of dialogue between patient and provider) (n=5). In only one encounter was the provider able to convince the patient that an opioid was not necessary. Table 4 describes communication features by whether or not conflict or disagreement was present in the discussion. Providers tended to more often express concerns about opioid therapy or suggest alternative treatments during encounters containing conflict. All but two of these discussions (n=10) ended with the provider changing the topic of conversation. Providers more frequently indicated high levels of frustration with patients during visits containing conflict (60% vs 42%), though this was not statistically significant.
Figure 2.
Outcome of Disagreements between Providers and Patients Regarding Opioid Analgesics
Table 4.
Communication Characteristics by Whether Conflict/Disagreement Occurred in Encounter
| Visits with Conflict | Visits without Conflict | p-value | |
|---|---|---|---|
| n=12 | n=36 | ||
| Provider expresses concern, n (%) | 9 (75.0%) | 13 (36.1%) | 0.019 |
| Provider suggests alternative therapy, n (%) | 10 (83.3%) | 13 (36.1%) | 0.005 |
| Provider acknowledges pain, n (%) | 6 (50.0%) | 14 (38.9%) | 0.499 |
| Provider changes topic, n (%) | 10 (83.3%) | 26 (72.2%) | 0.537 |
Although patients initiated most of the discussions in our sample, providers often ended the discussion by changing the topic (75%), sometimes abruptly (Table 3). Ultimately, providers more often agreed (50%) than disagreed (23%) to prescribe opioids, sometimes reluctantly (“I’ll give you a little bit of oxycodone but you need to get off that, okay?”). However, in 27% of encounters, no explicit decision to prescribe or not was made about the issue. Finally, when providers attempted to defer a pain management decision to another care provider, discussions sometimes became more antagonistic, and both parties became frustrated without an immediate solution to treat the pain. In some encounters (10%), providers agreed to a temporary prescription but deferred major decisions about pain management to other providers such as neurologists or orthopedic surgeons.
DISCUSSION
This study demonstrates that substantive discussion of opioid use for chronic pain management is common in ambulatory HIV practice and that these discussions are sometimes characterized by conflict, provider avoidance, and low provider empathy. Furthermore, we demonstrate that providers may have more negative attitudes towards patients with whom they discuss opioids, even after adjusting for possibly confounding social factors including illicit drug use. These findings together suggest that provider training in pain management and in communication skills specifically related to conflict management and negotiation might lead to improved quality of patient care and reduce provider frustration.
To our knowledge, this is the first study of its kind to use a large number of transcripts from actual patient visits and found that providers have lower regard for patients with whom they discuss opioids. These associations were not noted when analyzing visits in which opioids were briefly mentioned but not discussed. The fact that providers did not report poorer attitudes after these visits in which opioids were mentioned briefly suggests that lower regard may stem from these difficult conversations, rather than simply taking care of a patient in pain or on opioid therapy. While the associations in this study may not be causal, it is supported by prior work which examined provider narratives about their experiences with these patients. A study by Matthias and colleagues surveyed twenty providers about their experiences and found that many described the “emotional toll” of chronic pain care, “including feeling frustrated, ungratified, and guilty.”26 Providers face the difficult task of balancing the potential of patient abuse or diversion of prescription drugs with the possibility of under-treating the patient’s pain27. These concerns are particular salient for HIV-infected populations, where drug addiction is highly prevalent. Thus physicians must spend time negotiating with patients to adequately address their concerns while minimizing the potential for harm. Recognizing that opioids, particularly in high doses, may cause more harm than good28,29,30, it is imperative that providers are trained to effectively and safely manage pain with multiple modalities.
Given the difficult nature of these discussions, we identified some provider behaviors that could serve as targets for opioid-specific provider communication training. Some provider communication behaviors—such as avoiding the topic of opioids, or failing to acknowledge patients’ pain—may contribute to provider frustration during these visits. Interestingly, providers only acknowledged patient’s pain in about half of encounters, even in discussions containing conflict when patients tended to emphasize their pain symptoms. When providers attempted to avoid or shorten these conversations, they sometimes did not address important issues such as the underlying etiology of pain or patients’ admitted use of illegally obtained prescription opioids. Some providers attempted to diffuse the situation by deferring pain management decisions to another provider. While deferring management decisions may be unavoidable in some practice settings that require prescriptions to be written by pain specialists, providers might still convey to patients that treating pain is a priority, even if they are prohibited from prescribing opioids. Interestingly, we noticed that when provider and patients disagreed, the majority of providers eventually prescribed opioids, even after a long discussion. Ceding in the face of conflict likely contributes to providers’ feelings of frustration and low regard for patients10.
In contrast to the conflict and frustration we found in some encounters, we also noticed that some clinicians were extremely flexible and skilled in their discussions of opioids. Verbalizing concern for the patient and providing information about the risks associated with opioid therapy was a strategy used by many providers. Providers also suggested alternative treatments to show patients that treating pain was a priority, even if an opioid prescription could not be given. A recent study of audio-recorded visits with chronic pain patients identified additional helpful strategies for these discussions including reassurance, gathering additional information, and suggesting non-opioid medications12.
Pain management experts have suggested that provider empathy is an integral part of any encounter in which patients discuss pain31,32. Yet some authors have suggested that challenging encounters with these patients “can trigger a professional’s self-protective and defensive coping mechanisms that, in turn, can provoke decidedly unempathic responses.”33 When providers express empathy, they not only calm and comfort patients34, but also strengthen the patient-provider relationship by nurturing a collaborative atmosphere. Patients ultimately have more positive, productive relationships with providers who they view as being concerned about them and their suffering35. Even if providers feel manipulated by patients, or if they decide not to prescribe opioids against a patient’s wishes, the explicit acknowledgement of a patient’s pain and frustration has the potential to de-escalate a difficult situation. As we noted in our sample, providers can express empathy without needing to prescribe opioids. In fact, it is perhaps most important for providers to address patient suffering when not giving strong pain medication. The current study suggests that providers may underutilize empathy when discussing pain and opioid pain management with patients. Our findings inform the development of future interventions to improve provider communication: training providers to provide empathic feedback and focus on shared goal setting may result in a more patient-centered approach to pain management.
We acknowledge several limitations that should be considered when interpreting these findings. First, the twenty search terms used to identify visits containing relevant discussions may not have captured all pertinent dialogue in the study sample, and therefore our finding that 10% of visits included substantive discussion could be an underestimate. However, we did take a number of steps to ensure maximum sensitivity of our search including querying investigators from each study site, accounting for misspelling by transcriptionists, and using portions of search terms to account for uses of abbreviations or slang terms. Second, we sometimes found it difficult to infer the meaning of a particular excerpt of dialogue without having broader knowledge of the patient-provider relationship. To address this issue, we examined contextual dialogue before and after the discussion of opioids in an attempt to understand how these conversations related to the encounter as a whole. However, we do not have knowledge of what might have been discussed with respect to the pain and its therapy in other encounters. All of the encounters we examined were follow-up visits, and prior study has demonstrated that less time is spent discussing pain management when patients and providers are familiar with each other36. Analyses of first visits might reveal more discussion of pain and potentially more disagreement and conflict. In addition, we did not have data regarding which mediations were actually prescribed to patients, and thus we are unable to draw conclusions about the quantity or doses of medications prescribed. Finally, since our study population is drawn from HIV clinics at four sites across the country, this population might not be generalizable to the overall U.S. population in terms of demographic characteristics, the underlying amount of illicit drug use, provider training, frequency of visits, and nature of patient-provider relationships.
Our study was primarily descriptive and does not offer easy solutions to the well-known challenges of discussing opioid pain management. We believe, however, that unpacking and sorting the contents of these challenging discussions is a necessary step towards developing a better approach, and a better approach is clearly needed. Providers are often unable to communicate effectively with patients seeking opioid pain relief, tending to avoid discussion of the topic and failing to validate patients’ suffering. Providers can leave encounters frustrated, with negative attitudes towards these patients. Negative provider sentiments, coupled with the unproductive nature of these discussions, likely lead to lower quality care. Further research and attention to this particular communication challenge is needed to determine and teach optimal communication and negotiation skills around opioid treatment of chronic or severe pain. Improving the quality of these common discussions will not only better patient care, but might also reduce provider frustration and burnout.
Highlights.
Pain management discussions are common in outpatient HIV encounters.
Providers may regard patients less favorably if opioids are discussed.
These conversations were sometimes adversarial.
Empathy may facilitate more productive discussions.
ACKNOWLEDGEMENTS
Funders: This research was supported by a contract from the Health Resources Service Administration and the Agency for Healthcare Research and Quality (AHRQ 290-01-0012). In addition, Dr. Korthuis was supported by the National Institute of Drug Abuse (K23 DA019808), Dr. Beach was supported by the Agency for Healthcare Research and Quality (K08 HS013903-05), Dr. Saha was supported by the Department of Veterans Affairs, and both Drs. Beach and Saha were supported by Robert Wood Johnson Generalist Physician Faculty Scholars Awards. Dr. Hughes’s was supported by the Predoctoral Clinical Research Training Program at Johns Hopkins (UL1-RR025005)
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Prior presentations: These data were presented in part at the 34th Annual Meeting of the Society of General Internal Medicine in May 2011 (Phoenix, AZ); at the Communication, Medicine, and Ethics (COMET) Society Meeting in June 2011 (Nottingham, United Kingdom); and at the International Conference on Communication in Healthcare in October 2011 (Chicago, IL).
REFERENCES
- 1.American Pain Foundation. Pain facts and figures. [Accessed November, 2010]; Available at: http://www.painfoundation.org/page.asp?file=Newsroom/PainFacts.htm.
- 2.Caudill-Slosberg MA, Schwartz LA, Woloshin S. Office visits and analgesic prescriptions for musculoskeletal pain in US: 1980 vs. 2000. Pain. 2004;109(3):514–519. doi: 10.1016/j.pain.2004.03.006. [DOI] [PubMed] [Google Scholar]
- 3.Hing E, Hall MJ, Ashman JJ, Xu J. National Hospital Ambulatory Medical Care Survey: 2007 Outpatient Department Summary. [Accessed on November 5th 2010];National Health Statics Report. 2010 Sep 23; http://www.cdc.gov/nchs/data/nhsr/nhsr028.pdf. [PubMed]
- 4.Breitbart W, McDonald MV, Rosenfeld B, Passik SD, Hewitt D, Thaler H, Portenoy RK. Pain in Ambulatory AIDS Patients. I: Pain Characteristics and Medical Correlates. Pain. 1996 Dec;68(2–3):315–321. doi: 10.1016/s0304-3959(96)03215-0. [DOI] [PubMed] [Google Scholar]
- 5.Breitbart William, Rosenfeld Barry D, Passik Steven D, McDonald Margaret V, Thaler Howard, Portenoy Russell K. The Undertreatment of Pain in Ambulatory AIDS Patients. Pain. 1996 May;65(2–3):243–249. doi: 10.1016/0304-3959(95)00217-0. [DOI] [PubMed] [Google Scholar]
- 6.Rosenfeld Barry, Breitbart William, McDonald Margaret V, Passik Steven D, Thaler Howard, Portenoy Russell K. Pain in Ambulatory AIDS Patients. II: Impact of Pain on Psychological Functioning and Quality of Life. Pain. 1996 Dec;68(2–3):323–328. doi: 10.1016/s0304-3959(96)03220-4. [DOI] [PubMed] [Google Scholar]
- 7.Dobscha SK, Corson K, Flores JA, Tansill EC, Gerrity MS. Veterans Affairs primary care clinicians’ attitudes toward chronic pain and correlates of opioid prescribing rates. Pain Med. 2008;9:564–571. doi: 10.1111/j.1526-4637.2007.00330.x. [DOI] [PubMed] [Google Scholar]
- 8.Kenny DT. Constructions of chronic pain in doctor- patient relationships: Bridging the communication chasm. Patient Educ Couns. 2004;52:297–305. doi: 10.1016/S0738-3991(03)00105-8. [DOI] [PubMed] [Google Scholar]
- 9.Couto JE, Romney MC, Leider HL, Sharma S, Goldfarb NI. High rates of inappropriate drug use in the chronic pain population. Popl Health Manag. 2009 Aug;12(4):185–190. doi: 10.1089/pop.2009.0015. [DOI] [PubMed] [Google Scholar]
- 10.Eggly S, Tzelepis A. Relational Control in Difficult Physician–Patient Encounters: Negotiating Treatment for Pain. Journal of Health Communication. 2001;6:323–333. doi: 10.1080/108107301317140814. [DOI] [PubMed] [Google Scholar]
- 11.Werner A, Malterud K. It is hard work behaving as a credible patient: Encounters between women with chronic pain and their doctors. Soc Sci Med. 2003;57:1409–1419. doi: 10.1016/s0277-9536(02)00520-8. [DOI] [PubMed] [Google Scholar]
- 12.Matthias Marianne S, Krebs Erin E, Collins Linda A, Bergman Alicia A, Coffing Jessica, Bair Matthew J. ‘I’m Not Abusing or Anything’: Patient-Physician Communication about Opioid Treatment in Chronic Pain. Patient Education and Counseling. 2013 Nov;93(2):197–202. doi: 10.1016/j.pec.2013.06.021. [DOI] [PubMed] [Google Scholar]
- 13.Saha S, Sanders DS, Korthuis PT, Cohn JA, Sharp VL, Haidet P, Moore RD, Beach MC. The role of cultural distance between patient and provider in explaining racial/ethnic disparities in HIV care. Patient Educ Couns. 2011 Dec;85(3):e278–e284. doi: 10.1016/j.pec.2011.01.012. Epub 2011 Feb 18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Korthuis PT, Saha S, Chander G, McCarty D, Moore RD, Cohn JA, Sharp VL, Beach MC. Substance use and the quality of patient-provider communication in HIV clinics. AIDS Behav. 2011 May;15(4):832–841. doi: 10.1007/s10461-010-9779-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Beach MC, Saha S, Korthuis PT, Sharp V, Cohn J, Wilson IB, Eggly S, Cooper LA, Roter D, Sankar A, Moore R. Patient-provider communication differs for black compared to white HIV-infected patients. AIDS Behav. 2011 May;15(4):805–801. doi: 10.1007/s10461-009-9664-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Beach MC, Saha S, Korthuis PT, Sharp V, Cohn J, Wilson I, Eggly S, Cooper LA, Roter D, Sankar A, Moore R. Differences in patient-provider communication for Hispanic compared to non-Hispanic white patients in HIV care. J Gen Intern Med. 2010 Jul;25(7):682–687. doi: 10.1007/s11606-010-1310-4. Epub 2010 Mar 18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kumar R, Korthuis PT, Saha S, Chander G, Sharp V, Cohn J, Moore R, Beach MC. Decision-making role preferences among patients with HIV: associations with patient and provider characteristics and communication behaviors. J Gen Intern Med. 2010 Jun;25(6):517–523. doi: 10.1007/s11606-010-1275-3. Epub 2010 Feb 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kinsman H, Roter D, Berkenblit G, Saha S, Korthuis PT, Wilson I, Eggly S, Sankar A, Sharp V, Cohn J, Moore RD, Beach MC. "We'll do this together": the role of the first person plural in fostering partnership in patient-physician relationships. J Gen Intern Med. 2010 Mar;25(3):186–193. doi: 10.1007/s11606-009-1178-3. Epub 2009 Dec 22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale) Am J Prev Med. 1994;10:77–84. [PubMed] [Google Scholar]
- 20.Cacciola JS, Alterman AI, McLellan AT, Lin YT, Lynch KG. Initial evidence for the reliability and validity of a "Lite" version of the Addiction Severity Index. Drug Alcohol Depend. 2007;87:297–302. doi: 10.1016/j.drugalcdep.2006.09.002. [DOI] [PubMed] [Google Scholar]
- 21.Van Ryn M, Burke J. The effect of patient race and socio-economic status on physicians’ perceptions of patients. Soc Sci Med. 2000 Mar;50(6):813–828. doi: 10.1016/s0277-9536(99)00338-x. [DOI] [PubMed] [Google Scholar]
- 22.Hall JA, Horgan TG, Stein TS, Roter DL. Liking in the physician—patient relationship. Patient Educ Couns. 2002 Sep;48(1):69–77. doi: 10.1016/s0738-3991(02)00071-x. [DOI] [PubMed] [Google Scholar]
- 23.Beach MC, Roter DL, Wang NY, Duggan PS, Cooper LA. Are physicians' attitudes of respect accurately perceived by patients and associated with more positive communication behaviors? Patient Education and Counseling. 2006;62(3):347–354. doi: 10.1016/j.pec.2006.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Liang K, Zeger S. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73(1):13–22. [Google Scholar]
- 25.StataCorp. Stata Statistical Software: Release. Vol. 11. College Station, TX: StataCorp LP; 2009. [Google Scholar]
- 26.Matthias MS, Parpart AL, Nyland KA, Huffman MA, Stubbs DL, Sargent C, Bair MJ. The patient–provider relationship in chronic pain care: Providers’ perspectives. Pain Med. 2010;11:1688–1697. doi: 10.1111/j.1526-4637.2010.00980.x. [DOI] [PubMed] [Google Scholar]
- 27.Primm BJ, Perez L, Dennis GC, Benjamin L, Clark HW, Keough K, Leak WD, Payne R, Smith D, Sullivan LW. Managing Pain: The Challenge in Underserved Populations: Appropriate Use versus Abuse and Diversion. J Natl Med Assoc. 2004 Sep;96(9):1152–1161. [PMC free article] [PubMed] [Google Scholar]
- 28.Hay Justin L, White Jason M, Bochner Felix, Somogyi Andrew A, Semple Tim J, Rounsefell Bruce. Hyperalgesia in Opioid-Managed Chronic Pain and Opioid-Dependent Patients. The Journal of Pain. 2009 Mar;10(3):316–322. doi: 10.1016/j.jpain.2008.10.003. [DOI] [PubMed] [Google Scholar]
- 29.Katz MH. Long-Term Opioid Treatment of Nonmalignant Pain: A Believer Loses His Faith. Archives of Internal Medicine. 2010 Sep 13;170(16):1422–1424. doi: 10.1001/archinternmed.2010.335. [DOI] [PubMed] [Google Scholar]
- 30.Pud Dorit, Cohen Doron, Lawental Eli, Eisenberg Elon. Opioids and Abnormal Pain Perception: New Evidence from a Study of Chronic Opioid Addicts and Healthy Subjects. Drug and Alcohol Dependence. 2006 May 20;82(3):218–223. doi: 10.1016/j.drugalcdep.2005.09.007. [DOI] [PubMed] [Google Scholar]
- 31.Gallagher RM. Empathy: A timeless skill for the pain medicine toolbox. Pain Med. 2006;7:213–214. doi: 10.1111/j.1526-4637.2006.00163.x. [DOI] [PubMed] [Google Scholar]
- 32.Tait RC. Empathy: Necessary for effective pain management? Curr Pain Head Rep. 2008;12:108–112. doi: 10.1007/s11916-008-0021-6. [DOI] [PubMed] [Google Scholar]
- 33.Banja JD. Toward a more empathic relationship in pain medicine. Pain Med. 2008;9:1125–1129. doi: 10.1111/j.1526-4637.2008.00459.x. [DOI] [PubMed] [Google Scholar]
- 34.Banja J. Empathy in the physician’s pain practice: Benefits, barriers, and recommendations. Pain Med. 2006;7:265–275. doi: 10.1111/j.1526-4637.2006.00159.x. [DOI] [PubMed] [Google Scholar]
- 35.Upshur CC, Bacigalupe G, Luckmann R. “They don’t want anything to do with you”: Patient views of primary care management of chronic pain. Pain Med. 2010;11:1791–1798. doi: 10.1111/j.1526-4637.2010.00960.x. [DOI] [PubMed] [Google Scholar]
- 36.Henry SG, Eggly S. How Much Time Do Low-Income Patients and Primary Care Physicians Actually Spend Discussing Pain? A Direct Observation Study. J Gen Intern Med. 2011;27(7):787–793. doi: 10.1007/s11606-011-1960-x. [DOI] [PMC free article] [PubMed] [Google Scholar]