Correspondence to: Adriane N. Irwin, MS, PharmD, Oregon State University College of Pharmacy, 1601 SW Jefferson Way, Corvallis, OR 97331, USA. Tel: (541) 737-5722; Fax: (541) 737-3999; E-mail: adriane.irwin@oregonstate.edu.
Search for other works by this author on: Natalea Braden-Suchy, PharmD , Natalea Braden-Suchy, PharmD Oregon State University College of Pharmacy , Corvallis, Oregon, Search for other works by this author on: Daniel M Hartung, PharmD, MPH Daniel M Hartung, PharmD, MPH Oregon State University College of Pharmacy , Corvallis, Oregon, Search for other works by this author on: Pain Medicine, Volume 22, Issue 4, April 2021, Pages 970–978, https://doi.org/10.1093/pm/pnaa344 11 October 2020Adriane N Irwin, Natalea Braden-Suchy, Daniel M Hartung, Exploring Patient Perceptions of Opioid Treatment Agreements in a Community Health Center Environment, Pain Medicine, Volume 22, Issue 4, April 2021, Pages 970–978, https://doi.org/10.1093/pm/pnaa344
Navbar Search Filter Mobile Enter search term Search Navbar Search Filter Enter search term SearchDescribe patient attitudes toward opioid treatment agreements (OTAs) and characterize perceptions of their impact on patient care, behavior, communication with prescribers, and engagement with the health care system.
Qualitative descriptive study.
Federally qualified health center with six clinic locations serving a rural population.
Patients were prescribed long-term opioid therapy and were required to sign an OTA through an in-person office visit with a clinical pharmacist.
Patients who signed an OTA were recruited to participate in semistructured, in-person, one-on-one interviews. Data were analyzed using immersion-crystallization methods.
Among the 20 patients recruited, 50% were men; 70% were insured by the state’s Medicaid program; and 85% were using opioids for hip, back, and/or neck pain. Four major themes arose from the interviews. First, individuals who use long-term opioids experience a wide variety of opioid-related fears and stigmas. Second, individuals articulated real or potential benefits from implementing OTAs. Third, opinions differed on whether OTAs affected behavior and reduced opioid misuse and diversion. Finally, individuals provided feedback on the health care system’s OTA implementation process.
ConclusionsPatients experienced a wide variety of opioid-related fears and stigmas, including how OTA requirements can perpetuate these issues. Despite these feelings, participants articulated real or potential positive outcomes from the use of OTAs, although they were mixed on whether these agreements translated to any behavior changes. If OTAs are to become standard practice, future research is needed to describe the diversity of patient perspectives and experiences with OTAs and to evaluate their effect on patient outcomes.
The treatment of chronic pain, escalating use of opioids, and the public health epidemic posed by opioid use disorder (OUD) and unintentional overdose have been the focus of significant debate and public policy initiatives [ 1]. Physicians and other health care professionals have been placed in the predicament of balancing the need to provide patient-centered pain management with increased professional and regulatory scrutiny. To assist prescribers with navigating these decisions, multiple guidelines and consensus statements have been published that frequently recommend the use of risk reduction strategies, such as urine drug screens, prescription drug monitoring programs, and opioid treatment agreements (OTAs) [ 2].
OTAs, also called pain contracts, are frequently recommended as a strategy to increase safety and mitigate risk for opioid misuse and diversion [ 2]. These are written documents intended to outline expectations for opioid therapy by providing patient education, obtaining informed consent about the treatment course, and establishing monitoring terms [ 3]. In a review of 10 guidelines addressing opioid use for the treatment of chronic pain, four recommended the use of these agreements for all patients, with another five recommending use based on certain patient characteristics [ 2]. However, despite this widespread endorsement, little is known about the content of these agreements, the processes by which they are implemented, or their effectiveness in reducing outcomes related to OUD or diversion. Two systematic reviews—one published in 2010 and one published in 2020—have synthesized evidence on the effectiveness of OTAs in reducing and mitigating opioid misuse, with both concluding that evidence supporting use of these agreements on these outcomes is weak [ 4, 5].
Uncertainty regarding the effectiveness of OTAs in mitigating opioid misuse is not surprising. The pace of the opioid epidemic has far exceeded evidence development. As a result, despite weak evidence, proponents often cite other benefits, such as increases in provider-reported satisfaction, comfort, and mastery in managing chronic pain [ 6, 7]. However, unlike most uses for the term “agreement” or “contract” where conditions are agreed upon through negotiation by both parties, in this case, the terms of the OTA are generally nonnegotiable and patients are inherently positioned as the lower party in the patient-provider relationship by virtue of illness, knowledge base, and social hierarchy [ 8]. Many patients with chronic pain also have limited choices in health care providers and may feel coerced into signing these agreements for fear of jeopardizing relationships and not receiving necessary medical care [ 8]. Opponents argue that OTAs undermine the ethical principle of autonomy and jeopardize, rather than enhance, patient-provider relationships.
Unfortunately, many of these concerns are speculative because there is very limited research into patient perspectives of these agreements and the associated outcomes. Furthermore, interpretation of existing information is challenging, as it is derived from either patient surveys with conflicting outcomes or from focus groups that enrolled patients from a wide array of health systems and settings, including patients who signed OTAs in situations outside guideline recommendations (e.g., emergency departments) [ 9–11]. As a result, using patients who had experienced a similar care process, the purpose of this study was to address this literature gap by describing patient attitudes toward OTAs as well as characterize perceptions of the effect of OTAs on patient care, behavior, communication with prescribers, and engagement with the health care system.
This study was conducted with patients served by the Community Health Centers of Benton and Linn Counties, a federally qualified health center with six locations serving 8,000 patients in Oregon’s rural south Willamette Valley. The main health center is in Corvallis, OR, which has a population of 55,000, with smaller facilities in neighboring towns that range from 500 to 20,000 people. Four of the six health centers are Tier 5 Patient-Centered Primary Care Homes, a program administered by the Oregon Health Authority to formally recognize clinics that offer high-quality, patient-centered care by achieving criteria related to the following core attributes: access to care, accountability, comprehensiveness, continuity, coordination and integration, and patient and family centeredness [ 12]. Tier 5 is the highest designation [ 12]. Patients were recruited from two of the six health centers, specifically the health centers in Monroe and Lebanon, OR.
Through a collaboration with the Oregon State University (OSU) College of Pharmacy, pharmacy services are provided through a 340(B) community pharmacy, community-based pharmacy residency program, and clinical faculty embedded within the primary care teams. Clinical pharmacists provide face-to-face and telephone-based clinical services through collaborative practice agreements; respond to drug information requests; and assist with clinic initiatives focused on cost, access, and quality.
Clinical pharmacists are actively engaged in efforts to combat the inappropriate use of opioids and other controlled substances, including administering OTAs [ 13]. At all clinics, clinical pharmacists work directly with patients to review agreements during dedicated 60-minute face-to-face office visits. Organizationally, these agreements are referred to as “controlled substance agreements,” as they are signed for opioids as well as nonopioid controlled medications (e.g., benzodiazepines, stimulants). The organization's OTA includes a summary of terms and conditions (three pages), a worksheet to be completed by the patient on short- and long-term goals for pain management and options for the use of pharmacologic and nonpharmacologic therapies in achieving these goals (also known as shared decision-making) (one page), and a state-mandated risk notice for opioid therapy (two pages). Agreements are renewed every 2 years. Based on the preference of the patient’s primary care provider (PCP), pharmacists may also provide further support, such as interpreting urine drug screens, querying the Oregon Prescription Drug Monitoring Program, monitoring refill requests, and following up with patients on concerning behavior. All activities are documented in the electronic health record (Epic Corporation, Verona, WI).
This was a qualitative study of patients who had completed an appointment with a clinical pharmacist to review and sign an OTA. Patients were eligible for inclusion if they had signed an OTA in the 2 years prior to this study (i.e., the organization’s renewal period) for an opioid medication. Patients were excluded if they were younger than 18 years of age, had signed a controlled substance agreement for a nonopioid controlled medication, were pregnant, or were unable to provide written consent. Prior to study initiation, all study activities were reviewed and approved by the OSU Institutional Review Board.
A convenient sample was recruited through the following process. First, patients were identified through an administrative query of the electronic health record for those patients who had signed a controlled substance agreement in the last 2 years. Study recruitment began in July 2017, so this report encompassed the date range of June 15, 2015, to June 15, 2017. The administratively generated report was then manually reviewed to determine if the agreement was signed for an opioid medication, including tramadol, rather than a nonopioid controlled medication (i.e., benzodiazepines, stimulants) as well as any aforementioned exclusion criteria. Eligible patients were then recruited through a two-step process facilitated by two members of the research team who were also members of the organization’s primary care team(s) (Adriane N. Irwin and Natalea Braden-Suchy). In the first step, eligible patients were contacted by telephone by the clinical pharmacist assigned to the patient’s primary care team to provide a brief explanation of the study and to obtain consent to communicate the patient’s contact information to a study investigator. For this call, the clinical pharmacist communicated as a member of the patient’s care team. If permission was obtained, then a secure message was sent to the alternate researcher to have them contact the patient a second time. This call was still conducted by a clinical pharmacist, but it was a clinical pharmacist assigned to a different health center, and they communicated as a study investigator. During this second call, the patient was given a full explanation of the study and scheduled for a one-on-one interview if interested. Participants were offered an incentive of a $100 gift card. This amount was chosen because the health centers service large geographic areas, and this seemed appropriate to reduce barriers related to time and travel expenses. It was also consistent with a separate qualitative study focused on opioid safety occurring during approximately the same time period [ 14].
Recruitment efforts were ongoing from July to October 2017 and lasted until all eligible patients had received at least one outreach and at least 20 interviews had been collected. Recruitment was initially targeted for 20 patients (or 10 patients per participating clinic; i.e., health centers in Monroe and Lebanon, OR). This number was determined by reviewing other qualitative studies in the pharmacy literature in which the research question was believed to be of similar scope and the data were collected using similar methods. Additionally, we anticipated high-quality data, as participants would have experience with a similar agreement, workflow, and care team structure [ 15, 16]. There was flexibility to increase the sample size if needed based on interviewer judgment if it was believed that additional information was being gleaned from later interviews.
Data on participant characteristics were manually extracted from the electronic health record, whereas data for the qualitative analysis were collected through in-person, one-on-one interviews at the patient’s health center in a private space. All interviews were conducted by clinical pharmacists within the health system who were not directly involved with the patient’s care (Adriane N. Irwin and Natalea Braden-Suchy). As a result, both interviewers had direct experience with the organization’s OTA as well as foundational knowledge of clinic workflows and care team structure.
These two interviewers collaboratively developed and pilot-tested a semistructured interview guide to provide consistency between interviews ( Table 1). Participant interviews generally lasted 30 to 60 minutes; the interview included these standardized questions and then probing questions as needed to clarify participants’ answers. Written consent was obtained from all participants prior to the interview.
Patient interview script
Patient interview script
Each interview was audio-recorded and then transcribed verbatim by a professional transcription company. Transcripts were not returned to participants for comment or correction prior to analysis. Qualitative analysis was then performed to identify emerging themes using an immersion-crystallization approach [ 17]. Two study investigators (Adriane N. Irwin and Natalea Braden-Suchy) plus two experts in qualitative analysis methods (Lindsey Alley and Jennifer Hall) to ensure confirmability met to review transcripts line by line (read data aloud), discuss identified concepts, and develop the initial coding structure using a sample of the transcripts. This initial coding structure was then applied to another sample of transcripts; two study investigators (Adriane N. Irwin and Natalea Braden-Suchy) independently reviewed the transcripts to pilot-test the coding dictionary and assess consistency between coders. After the pilot test, the full group reconvened to discuss and revise the coding structure before applying it to the remaining interviews. Coding was again conducted by the two study investigators (Adriane N. Irwin and Natalea Braden-Suchy). Regular meetings were used to allow investigators to discuss codes, emerging themes, and gaps or inconsistencies in the analysis. No new codes were added in the coding of the later interviews, indicating that the coding structure captured all themes and that data saturation had been achieved. Coded text was then categorized, and analytic memos were written and then discussed to identify the first three themes. After the first three themes were developed, the data were reviewed again to assess best-practice recommendations, which then resulted in the final theme. During this process, meeting frequency ranged from multiple times per week in the beginning to weekly and then monthly as study investigators coded larger blocks of text independently or drafted analytic memos.
Data on participant characteristics were summarized with descriptive statistics, with nominal data presented as a number (percentage) and continuous data presented as the mean ± standard deviation. Quantitative analysis was performed with Excel (Microsoft, Redmond, WA, USA). Qualitative data were organized and coded using NVivo version 10 (QSR International, Melbourne, Australia). All data storage as well as file transfers for the participant interviews were conducted using a secure cloud content management and file-sharing platform (Box, Inc., Redwood City, CA, USA).
A total of 20 interviews were completed. Table 2 lists the participant characteristics. Participants were evenly divided between men and women (n=10; 50%), with an average age of 52.8±9.9 years, and had been established with the clinic for an average of 3.9±2.0 years. Most were covered by the state’s Medicaid program (n=14; 70%). They were also most frequently using chronic opioids for hip, back, and neck pain (n=17; 85%) with a morphine equivalent dose of 42.0±31.2 mg daily.
| Characteristic . | Patients (n = 20) . |
|---|---|
| Age, years ± SD | 52.8 ± 9.9 |
| Men, n (%) | 10 (50) |
| Payer, n (%) | |
| Medicare | 4 (20) |
| Medicaid | 14 (70) |
| Private | 2 (10) |
| Years with clinic, years ± SD | 3.9 ± 2.0 |
| Medications, number ± SD | 11.6 ± 5.1 |
| Opioid indication(s), * n (%) | |
| Hip, back, and neck pain | 17 (85) |
| Other joint pain † | 7 (35) |
| Miscellaneous † | 2 (10) |
| Morphine equivalent dose, ‡ mg ± SD | 42.0 ± 31.2 |
| Characteristic . | Patients (n = 20) . |
|---|---|
| Age, years ± SD | 52.8 ± 9.9 |
| Men, n (%) | 10 (50) |
| Payer, n (%) | |
| Medicare | 4 (20) |
| Medicaid | 14 (70) |
| Private | 2 (10) |
| Years with clinic, years ± SD | 3.9 ± 2.0 |
| Medications, number ± SD | 11.6 ± 5.1 |
| Opioid indication(s), * n (%) | |
| Hip, back, and neck pain | 17 (85) |
| Other joint pain † | 7 (35) |
| Miscellaneous † | 2 (10) |
| Morphine equivalent dose, ‡ mg ± SD | 42.0 ± 31.2 |
SD = standard deviation.
Categories are not mutually exclusive.
Joint pain included pain of the knee, shoulder, or hand. Miscellaneous indications were fibromyalgia and headaches.
Morphine equivalent dose at the time of the interview.
| Characteristic . | Patients (n = 20) . |
|---|---|
| Age, years ± SD | 52.8 ± 9.9 |
| Men, n (%) | 10 (50) |
| Payer, n (%) | |
| Medicare | 4 (20) |
| Medicaid | 14 (70) |
| Private | 2 (10) |
| Years with clinic, years ± SD | 3.9 ± 2.0 |
| Medications, number ± SD | 11.6 ± 5.1 |
| Opioid indication(s), * n (%) | |
| Hip, back, and neck pain | 17 (85) |
| Other joint pain † | 7 (35) |
| Miscellaneous † | 2 (10) |
| Morphine equivalent dose, ‡ mg ± SD | 42.0 ± 31.2 |
| Characteristic . | Patients (n = 20) . |
|---|---|
| Age, years ± SD | 52.8 ± 9.9 |
| Men, n (%) | 10 (50) |
| Payer, n (%) | |
| Medicare | 4 (20) |
| Medicaid | 14 (70) |
| Private | 2 (10) |
| Years with clinic, years ± SD | 3.9 ± 2.0 |
| Medications, number ± SD | 11.6 ± 5.1 |
| Opioid indication(s), * n (%) | |
| Hip, back, and neck pain | 17 (85) |
| Other joint pain † | 7 (35) |
| Miscellaneous † | 2 (10) |
| Morphine equivalent dose, ‡ mg ± SD | 42.0 ± 31.2 |
SD = standard deviation.
Categories are not mutually exclusive.
Joint pain included pain of the knee, shoulder, or hand. Miscellaneous indications were fibromyalgia and headaches.
Morphine equivalent dose at the time of the interview.
Four themes arose from the participant interviews. First, individuals who use long-term opioids experience a wide variety of opioid-related fears and stigmas. Second, individuals articulated real or potential benefits from implementing OTAs. Third, opinions differed on whether OTAs affected behavior and reduced opioid misuse and diversion. Finally, individuals provided feedback on the health system’s OTA implementation process.
Participants expressed a diversity of opioid-related fears. Fear of developing an OUD—either themselves or by those around them—was frequently cited. However, these specific fears were highly diverse and reflected effects at the individual (e.g., fear of addiction or having medication stolen), community (e.g., fear of increased crime throughout the community), and government (e.g., fear of the impact of state policy or oversight on opioid access) levels.
“I don’t want to get to the point of waking up and that’s the first thing that I think of . . . I don’t want to be that person.” (M7)
More specific to OTAs, participants primarily discussed fears of losing access to opioids and the effect this would have on their ability to function. All participants reported knowing someone who had used opioids that were not prescribed to them or in a manner that was inconsistent with the prescribed directions, with many having experience with OUD in some capacity. Although many expressed empathy toward those individuals struggling with OUD, comments were frequently paired with discussion of how opioid misuse and diversion by others has negatively affected their provider’s willingness to prescribe opioids. Some felt judged by the actions of others, which contributed to their anxiety and, in a few cases, anger.
“I know of people that are not taking them. They are just selling them and stuff. And then it makes it bad for people like me or some in my family who have to be on them. It’s really bad because it doesn’t count the few that are doing it the right way.” (M8)
“. . . it affects me by putting me on a pain contract and making me feel like a criminal when I don’t have an issue.” (EL8)
There was discussion of how the initial request to sign an OTA triggered feelings of fear and anxiety, although these feelings were often alleviated after receiving more information and then completing the appointment. Additionally, there was discussion of how certain requirements (e.g., urine drug testing) or simply having an OTA on file translated to stigma and a perception that their care may be negatively affected in other places.
“I understand that the ER is protecting themselves, but they don’t need to look at the person like, well, you’re on it because you are a bad person. They don’t know me or why I’m on it.” (EL9)
Finally, for those who previously had an OTA with another medical office, some reported being dismissed due to nonadherence with the agreement or feeling pushed out because they were on opioids. They described policy changes where medical offices were no longer willing to continue prescribing at past amounts—or sometimes any amount—even if the patient reported being stable on those opioid regimens.
“. . . clinic told me that they are no longer honoring pain contracts or dealing with opioids and in essence fired me, giving me 30 days’ supply and that was it.” (M3)
Participants were all aware of the opioid crisis and the broader societal factors that led to the decision to implement OTAs by the health system. Not all participants agreed with this requirement, but everyone could discuss real or potential benefits to the patient, their PCP, or the health system.
Most consistently, participants discussed how these appointments enhanced communication between the patient and their care team. Participants stated that these appointments were educational and that they felt better informed on the opioid crisis and the risks of using prescription opioids. In some cases, this translated to behavioral changes aimed at improving safe opioid use.
“It may have brought [opioid-related risks] more to my attention. . . . It made me research them more. And I’m more aware of people that abuse, or people that are addicted, or people that are in trouble.” (EL3)
Participants also discussed how these appointments helped set expectations related to opioid use and increased patient accountability. Some viewed this focus on accountability favorably because it provided some protection for their opioid access, as it provided a vehicle for them to demonstrate that they could act responsibility.
“It makes you just really mind your P’s and Q’s. If you want to continue to have your pain pills, then you mind your P’s and Q’s.” (M1)
Additionally, participants discussed how these appointments provided an opportunity for shared decision-making on the treatment plan, including the role of opioids. However, opinions were mixed on whether patients actually had a voice in these decisions as implemented.
Finally, from the prescriber or health system lens, participants discussed how these agreements might provide liability protection if a negative outcome were to occur. Some participants also blamed prescribers for the opioid epidemic. They discussed how these agreements might help curb overprescribing and cause prescribers to reflect more on their opioid-prescribing decisions.
Although there was some consensus on the need for OTAs, opinions clearly diverged when the discussion shifted to whether these agreements affected behavior. For those who did feel that OTAs affected behavior, their rationale fell into two categories. First, this behavior change was caused by an increased awareness of opioid-related risks and the need to take opioid safety seriously.
“. . . and it’s made me treat it more like a controlled substance. Instead of just, you know, throwing it in my pocket or my car, I keep it locked up.” (M2)
Second, this behavior change was linked to a specific aspect of the agreement, such as the policy prohibiting early refills or a fear that OTA nonadherence would result in termination of opioid prescribing. The enforcement of these agreements was largely viewed as instrumental in their success.
“I think it’s more than just paperwork because it actually makes people more self-aware . . . not expecting that if you run out [of] them, you are going to be able to get more, because it’s not going to happen.” (EL7)
However, other participants felt that these agreements had little impact on patient behavior. This again diverged into two categories. Some stated that they would use prescription opioids as directed regardless of the OTA requirement. This is because their behavior stems from their personal beliefs and ethics and not from the agreement. Others viewed the agreements merely as paperwork because everyone will agree to the outlined terms and conditions . Those with a legitimate medical need for opioids will “agree” because the OTA is required for opioid access, and those with an OUD or other motivation (e.g., diversion) will “agree” because they have no intention of following the outlined terms.
“Some don’t even listen and just agree . . . they’ll just say it so they can get it.” (EL10)
Participants provided feedback on the process by which each clinic implemented the OTA requirement. This feedback was largely focused in three areas: how patients were notified of this requirement, the content of the agreement, and the in-person appointment.
The manner in which this requirement was communicated to participants clearly affected feelings toward the appointment. Most specifically, individuals who were notified by telephone call, generally from a medical assistant or customer service agent, expressed a trigger of fear and anxiety toward this appointment, whereas those informed by their PCP beforehand did not recall a similar trepidation during the preappointment period. Those with initial fear and anxiety, however, did report that these feelings often subsided upon completing the appointment, and the overall feedback about the appointments was generally positive.
“. . . and caught me off guard at first. Like I said, I was a little nervous. But, once I went in and she started talking, I realized that, oh, this isn’t, you know, the big scary . . . you know, some scary contract. It’s like any time the word ‘contract’ is used, it’s just like, what am I doing?” (EL1)
The communication style of the person conducting the appointment was important to many participants. Participants expressed a desire for openness, professionalism, and communication at their level. They also appreciated the opportunity to review the OTA with someone one-on one rather than it being part of registration or other check-in paperwork.
“Her explanation of what was going on. She answered my questions professionally and I was able to understand.” (EL7)
“Well, okay, I have to meet with a pharmacist. She’s going to know about what is going on. It’s not going to be somebody that is like a secretary or something that’s sitting behind a desk and they just give me the paper and I sign it and away I go.” (M2)
Thoughts were specifically solicited about the use of a clinical pharmacist to complete these agreements rather than the patient’s PCP. Most expressed no preference for who reviewed the agreement as long as the individual was part of the care team and was educated on opioids. For those who did have a preference, some did express a preference for their PCP. Conversely, some expressed appreciation for a different team member, as it provided more perspectives on the situation and opportunities for education on opioid use.
“It feels like it should be a neutral third party compared to maybe like a give-or-take between a doctor and patient . . . it is more emotionally removed from the doctor-patient relationship . . . just because I feel like people would be willing to ask more questions or feel more comfortable presenting concerns to a third party rather than directly to their doctor.” (EL5)
All participants were aware that they had signed an OTA, and many were able to cite basic themes; however, none were able to discuss the content in great detail outside a few specific terms.
“I’m hoping that you are not going to ask me what the contract said because I don’t remember exactly. In general, I remember the contract as being things like—you can’t ask for a prescription early, you can’t call in and say that you lost your prescription. . . .” (EL6)
Finally, participants consistently expressed appreciation for their PCP as well as the full care team. None of the participants felt that the request to sign an OTA and associated processes had negatively affected their care or undermined their overall positive feelings toward the clinic and their health care team.
“. . . just because we had that open communication. The key is always that open communication. If I had issues, I could call and make an appointment, and we’d talk and we’d come to an agreement.” (EL9)
The use of OTAs is increasingly viewed as a best practice for opioid prescribing, despite weak evidence supporting effectiveness in reducing opioid misuse and diversion [ 4, 5]. Although proponents contend that these agreements promote safety by setting expectations and increasing the likelihood that patients will adhere to a plan of care, opponents argue that these agreements undermine the ethical principle of autonomy and jeopardize patient-provider relationships [ 8]. In our research of patients prescribed long-term opioids, we found evidence supporting both positions. Although opinions were mixed on the ability of these agreements to alter behavior—a conclusion consistent with other qualitative research of patients and providers [ 11]—participants consistently cited real or potential positive benefits. These benefits frequently centered on improved patient education and communication, which are foundational elements for the endorsement of OTAs. However, patients also described a variety of opioid-related fears and stigmas, including how OTAs perpetuated these concerns. This included discussion of how these agreements felt coercive and made patients feel marginalized.
Existing data on patient perceptions of OTAs are limited. Two studies have attempted to measure patient perceptions after implementation of an agreement using a patient survey, with mixed results. In one study, more than 90% of patients agreed that the use of a patient-provider agreement was helpful in informing the course of therapy, and researchers found no evidence that this agreement adversely affected patient-physician relationships, although it is unclear if the consequences for OTA nonadherence were discussed [ 9]. However, in a different study, three-quarters of patients treated in a family medicine practice reported negative attitudes toward the clinic’s process, the physician, and/or a specific term of the agreement [ 10]. Additionally, one study has attempted to explore this topic through qualitative research via focus groups [ 11]. However, the applicability of the findings is challenging because most patients were asked to sign an agreement during emergency department or presurgical visits. These are settings that are inconsistent with the rationale for how OTAs should be used to improve patient care, and these are settings that are likely affected by duress and power imbalances. As a result, when taken together, it appears that these agreements have the potential to be powerful tools in improving opioid safety, but they need to be implemented thoughtfully to meet the needs of both patients and providers.
The importance of clear communication in the implementation of these agreements arose throughout the interviews, which also encouraged comparisons to existing work in this space. Research into agreements used by major academic pain centers or those that are publicly available from private-practice physician websites has shown significant diversity in decision-relevant content [ 18, 19]. Similarly, another study focused on literacy demands and formatting characteristics reported that most agreements were written at or above the 14th reading grade (college) level and contained both sophisticated medical and multisyllable nonmedical vocabulary not used in typical everyday conversation [ 20]. This has raised concerns about whether these OTAs truly enhance decisions surrounding opioid prescribing. Stakeholders, including a work group convened by the US Food and Drug Administration (FDA), have responded by developing agreement templates to facilitate standardization [ 21–23]. Research into the template developed by the FDA work group has reported that patients have responded favorably and do feel better informed on the course of therapy [ 9].
It should be noted that the template developed by the FDA work group was administered in less than 10 minutes [ 9]. This starkly contrasts with the 60-minute dedicated appointments used by our health system for these discussions and with the strengths cited by our participants—specifically comments on how participants did not feel rushed and felt that their questions were thoroughly answered at a level they could understand. Our research was conducted with an underserved population, and the time dedicated to these conversations may not be necessary (or desired) for other patient populations. However, patient recall of specific terms within these agreements has been reported to be poor, and the time required for patient questions has been raised as a substantial barrier to implementation, suggesting the need for more research into best practices for implementation [ 10, 11].
This study has several limitations. First, it was limited to patients within a single health system. This provided some strengths because all participants had experience with a similar agreement, workflow, and care team structure, which allowed for collection of high-quality data and the ability to reach data saturation with fewer interviews than is often required for qualitative research involving semistructured interviews [ 24]. However, the policies and workflows implemented by other health systems related to OTAs may evoke different perceptions. Similarly, all participants were recruited from a rural area, and most were insured by the state’s Medicaid program. As a result, patient perceptions may not be consistent with the perceptions of urban residents or those who are more affluent. Second, there were challenges with recruitment. Despite the first outreach coming from an individual with an existing relationship with the patient, some patients declined to participate at both steps. Additionally, some patients could not be reached at both steps due to unstable living situations or contact information. As a result, it is unclear if those willing and able to participate reflected the larger population even within our health system. Finally, the results are limited by the truthfulness of the patients. As shown through these interviews, opioid access is an emotional topic for many patients. In addition, these interviews intersected with areas where patients are less likely to be truthful, such as OUD, OTA nonadherence, and criminal activity. Member checking (e.g., interview transcript review) is frequently employed to increase data validity. We did not use this technique due to a low likelihood for participant engagement due to communication challenges—secondary to both sociodemographic characteristics (e.g., unstable phone numbers) and the rural location of the health centers (e.g., poor broadband coverage, travel distances)—as well as concerns about the literacy levels of some participants that would negatively affect their ability to complete this request. There were also concerns that interview transcript review could potentially reduce accuracy if patients chose to remove sensitive comments, and there were comments about behaviors suggestive of OUD and criminal activity. Throughout the study, we took significant steps to make patients feel comfortable and ensure confidentiality to develop trust. Because both interviewers had direct experience with the agreement and knowledge of the health center and surrounding community, they were able to ask informed follow-up questions to help address validity concerns. As a result, we believe that most participants were truthful, although it is not guaranteed, and the data analysis is based on what was stated during the interview.
In our qualitative analysis of patients prescribed long-term opioids, patients experienced a diversity of opioid-related fears and stigmas, including how OTA requirements can perpetuate these issues. Despite these feelings, participants articulated positive outcomes from the use of OTAs and described enhanced patient knowledge of the risks of opioid use and improved communication between the patient and care team, although participants were mixed on whether these agreements translated to any behavior changes. If OTAs are to become standard practice, then future research is needed to describe the diversity of patient perspectives and experiences with OTAs and to evaluate their effect on patient outcomes.
Funding sources: This work was partially funded by a New Investigator Award from the American Association of Colleges of Pharmacy.
Conflicts of interest: There are no conflicts of interest to report.
Past presentation: Portions of this research were presented at the 2018 Annual Meeting of the American Association of Colleges of Pharmacy in Boston, MA, on July 22, 2018.
The authors would like to thank Lindsey Alley, MS, at the Oregon Health & Science University (OHSU) Oregon Institute of Occupational Health Sciences and Jennifer Hall, MPH, at the OHSU School of Medicine for their assistance with the qualitative analysis.