Confronting challenges to opioid risk mitigation in the U.S. health system: Recommendations from a panel of national experts

Background Amid the ongoing U.S. opioid crisis, achieving safe and effective chronic pain management while reducing opioid-related morbidity and mortality is likely to require multi-level efforts across health systems, including the Military Health System (MHS), Department of Veterans Affairs (VA), and civilian sectors. Objective We conducted a series of qualitative panel discussions with national experts to identify core challenges and elicit recommendations toward improving the safety of opioid prescribing in the U.S. Design We invited national experts to participate in qualitative panel discussions regarding challenges in opioid risk mitigation and how best to support providers in delivery of safe and effective opioid prescribing across MHS, VA, and civilian health systems. Participants Eighteen experts representing primary care, emergency medicine, psychology, pharmacy, and public health/policy participated. Approach Six qualitative panel discussions were conducted via teleconference with experts. Transcripts were coded using team-based qualitative content analysis to identify key challenges and recommendations in opioid risk mitigation. Key results Panelists provided insight into challenges across multiple levels of the U.S. health system, including the technical complexity of treating chronic pain, the fraught national climate around opioids, the need to integrate surveillance data across a fragmented U.S. health system, a lack of access to non-pharmacological options for chronic pain care, and difficulties in provider and patient communication. Participating experts identified recommendations for multi-level change efforts spanning policy, research, education, and the organization of healthcare delivery. Conclusions Reducing opioid risk while ensuring safe and effective pain management, according to participating experts, is likely to require multi-level efforts spanning military, veteran, and civilian health systems. Efforts to implement risk mitigation strategies at the patient level should be accompanied by efforts to increase education for patients and providers, increase access to non-pharmacological pain care, and support use of existing clinical decision support, including state-level prescription drug monitoring programs.


Introduction
Over the past twenty-five years, opioid analgesics have been commonly prescribed in the U.S. for acute and chronic pain management [1]. In response to growing evidence that use of opioids for management of chronic pain may not be effective and is associated with significant risk of developing opioid use disorder, death by overdose, and other health conditions [1][2][3][4], U.S. healthcare and governmental organizations have taken a variety of steps to reduce opioid prescribing and overdose risk. Examples include implementing new policies for opioid prescribing in Massachusetts [5], disseminating clinical practice guidelines (e.g., CDC Guidelines for Prescribing Opioids for Chronic Pain) [6,7], and using prescription drug monitoring programs (PDMPs) [8] and other clinical decision support (CDS) to aid in risk assessment and management [9,10].
Although overall rates of opioid prescribing have declined since 2011 [11], opioid-related overdose deaths continue to increase suggesting that simply reducing opioid use in clinical care may not reduce opioid-related mortality [12]. In 2017 alone, 47,600 persons died in the U. S. from overdoses involving opioids, accounting for 67.8% of all drug overdose deaths [12]. For this and other reasons, the debate continues over whether there is an appropriate role for opioid medications within chronic pain management [13]. Efforts to reduce opioid-related harms face the additional challenge of healthcare fragmentation, with the Military Health System (MHS), Department of Veterans Affairs (VA), and public and private civilian health systems operating within unique organizational structures independently treating specialized but frequently overlapping patient populations. For example, the patient populations for military, VA, and civilian healthcare systems are in many ways distinct. Service members are on average considerably younger than Veterans, a higher preponderance of men receive care in military and VA settings, and individuals relying on VA health care tend to have more health conditions than those who do not [14][15][16]. At the same time, there is also frequent overlap and intersection between the patient populations served by these systems, particularly when considering that many Veterans receive some or most of their care in military or civilian treatment facilities, and many of those receiving care in military treatment facilities are non-Veteran dependents or retirees who may also be receiving care from outside the military system. Relatively little of the health and policy literature has been devoted to considering the unique and shared challenges faced by the separate healthcare systems operating within the U.S. [17], or the implications of their differing models of pain care for individuals receiving care across these systems [18].
In sum, achieving safe and effective pain management that includes opioids while mitigating the risks associated with opioids remains a significant challenge across the fragmented U.S. healthcare system. There is a need to identify core challenges and develop a focused agenda for safe and appropriate access to opioids within and across diverse healthcare systems. To address this gap, we conducted a qualitative study with a panel of national experts. Our objectives were to: (a) identify central challenges in opioid prescribing and risk mitigation; and (b) determine how best to support providers in operationalizing safe and appropriate opioid prescribing.

Methods
Semi-structured in-depth qualitative panel discussions were conducted with healthcare providers, pain and substance use clinical and health services researchers, and individuals with a policy-making or other operational leadership role (henceforth: policy makers) working within MHS, VA, and civilian environments. Expert panel discussions were selected as an effective method for learning from participants with expertise in a highly-specialized topic area [19,20]. Expert panel discussions are recommended as a qualitative method when existing scientific evidence is complex (e.g., mixed for/against a specific therapy) and/or requires substantial synthesis [19,21]. Given that opioid prescribing in the U.S. occurs across multiple healthcare systems and clinical settings (e.g., primary care, emergency department), panel discussions were determined to be an appropriate method for gathering and synthesis of diverse perspectives. Panel discussions were conducted as part of a larger evaluation of novel opioid risk mitigation tools to identify unhealthy opioid use and unsafe clinician prescribing patterns in the MHS [10].

Data collection strategies
We asked each participant to attend two separate expert panels in an effort to ensure sufficient opportunity for in-depth discussion. In order to facilitate participation by busy experts, we offered group discussions at six different times. Fourteen participants were able to attend two of the scheduled sessions; the remaining four experts were accommodated in individual sessions, for a total of 18 participants. Expert panel discussions were conducted via Adobe Connect in May 2017, and led by a PhD-level anthropologist or clinical psychologist with expertise in qualitative methodology. A semi-structured discussion guide was developed to elucidate perspectives on: (a) challenges in opioid risk mitigation (sample question: "what do you see as the biggest challenges for providers in prescribing opioids as part of their practice?"); and (b) how best to support providers in safe and appropriate opioid prescribing across MHS, VA, and civilian health systems (sample question: "what do providers need to ensure safe and appropriate opioid prescribing for their patients?"). Participants were informed that discussions were confidential and would be recorded (audio/visual); verbal consent was documented as part of recorded discussions. The UT Health San Antonio Institutional Review Board approved all study procedures.

Analysis
A team-based qualitative content analysis approach was used, with key themes emerging and refined during review, synthesis, and coding of discussion transcripts [22]. Panel and individual discussions were transcribed verbatim by a professional service (GMR Transcription) and uploaded to ATLAS.ti 7.0 [23] qualitative software. Three team members reviewed transcripts to identify initial themes related to challenges and recommendations. A coding manual was developed for the purpose of assigning categories (codes) to text reflecting these themes, and refined as coding proceeded. Two members of the research team (MC, SS) conducted independent coding of all transcripts, meeting weekly with a third member (EF) to identify coding disagreements, which were resolved by discussion to consensus. Individual themes and larger content areas (domains) were refined in team discussions. Themes were assigned to domains as appropriate. Following coding, the content of coded text was reviewed and synthesized. All team members participated in refining the final summation of themes.

Challenges
Challenges in chronic pain management and opioid risk mitigation fell within five overarching domains: chronic pain-related challenges; sociopolitical climate; health systems; providers; and patients (summarized in Table 2).
Chronic pain-related challenges. Chronic pain management is technically complex. Panel participants noted that delivery of effective care for chronic pain requires in-depth knowledge of patient assessment, appropriate medications and dosing, potential drug interactions, behavioral health and other non-pharmacological treatments, and recommended guidelines for care. They acknowledged that delivering these treatments in real time could pose a challenge.
There's this reality that all the decisions that go forward require a very individualized, patient-centered aspect, and we have trouble operationalizing that. We're trying, obviously, but it's tough. [ There is a lack of consensus around evidence-based strategies for effective chronic pain management and opioid risk mitigation. Panelists cited a lack of evidence regarding which treatments are safe and effective for chronic pain, what types of patients and conditions are most

PLOS ONE
Expert panel recommendations for opioid risk mitigation The "opioid crisis" has created a climate of tension around opioid prescribing.
So we have to look at providers because they are stuck in this quagmire of media frenzy and this challenge to reduce opioid use. [Civilian and VA Clinician and Researcher] Providers feel pressure to rapidly reduce the number and dose of opioid prescriptions for patients with chronic pain.
There's a lot of pushback against physicians from their leadership and also just from, uh, they-our state organizations, national organizations to avoid using these things, and it's not worth losing your livelihood. . .

[Civilian Clinician]
There is significant concern about negative consequences of rapid reduction in opioid availability.
I'm always concerned because I know there are some providers who are going to respond to this in a way that is not so good for patients, who are going to say, "I'm just going to stop prescribing, everybody is going to be tapered to this dose," and those sorts of things. [Civilian Clinician and Policy]

Poor interconnectivity between electronic medical records and other data monitoring systems results in gaps in data and creates burden for providers.
There is a push here. . . to make it mandatory, to query the PDMP for all patients before you prescribe an opioid or administer an opioid. And there is push back from various advocacy groups . . .because they feel like it takes too much time.

[Military Clinician]
Chronic pain patients may enter a system through a variety of specialties, raising challenges for care coordination, follow-up, and identification of potential risk or misuse.
If the clinician is looking at a prescription pad and there's not a mixture of physical therapy and maybe a pain psychologist and a bona fide mental-health provider for the patient's incipient PTSD, then that-all of that becomes very high-risk situation. [VA Clinician and Researcher] (Continued ) likely to benefit from opioids or other treatments, and ambiguity in how best to measure treatment benefits. One panelist stated that, "We don't have a good way to treat pain." Another panelist cited evidence that a patient's prior mental health and substance use history can serve as an important signal for opioid misuse potential; however, the same expert noted the evidence is not entirely clear and may serve as a "green light for [offering opioids to] the low-risk patient." Panelists also noted that, while there is significant work ongoing to develop tools and strategies in support of safe and appropriate chronic pain management, many of the existing strategies lack a strong evidence base.
The strategies are put out there to reduce risks-they are risk stratification and monitoring so you. Sociopolitical climate as a challenge. At the national level, the "opioid crisis" has created a climate of tension around opioid prescribing. Panelists described a national climate in which policy discussions about effective strategies for managing the crisis may be shaped by the emotional tenor of public debate, and in which even informational tools and resources may be "weaponized" and used to penalize patients and/or providers. They delineated a history of rapid change in the expected standard of care for patients with chronic pain, detailing a shift from heavy reliance on long-term opioids to significant concern about any use of opioids on a long-term basis.
So I think the biggest challenge is a lot of the political rhetoric around opioid use. No question opioids are dangerous but I think it has gotten way blown out of proportion compared to other medication, compared to other issues. NSAIDS for example have a very high risk of death from a coronary heart disease, hospitalizations, kidney dysfunctions, dialyses. Not to minimize the risk with opioids but we don't want to forget many other drugs have issues.

[Civilian and VA Clinician and Research]
Providers feel pressured to rapidly reduce the number and dose of opioid prescriptions for patients with chronic pain. In response to concerns about opioid prescribing, there has been a rapid swing within many healthcare organizations toward reducing the number of new prescriptions, and toward reducing the number and dose of long-term opioid prescriptions among existing patients with chronic pain. Providers may end up feeling "stuck in this quagmire of media frenzy and this challenge to reduce opioid use," with implications for how they approach prescribing: I'm always concerned because I know there are some providers who are going to respond to this in a way that is not so good for patients, who are going to say, I'm just going to stop prescribing, everybody is going to be tapered to this dose, and those sorts of things.

[Civilian Clinician and Policymaker]
There is significant concern about negative consequences of rapid reduction in opioid availability. Panelists described fears that the rapid shift in prescribing practice was resulting in a lack of patient-centered care and potential harm for patients. Panelists working across sectors perceived significant danger in failing to provide effective treatment for patients who had received long-term opioids for chronic pain, in particular describing the risks of leaving patients to "search for solutions on their own," with the potential for opioid misuse, use of illegal substances, addiction, overdose, or suicide.
I've observed a real rush to make the opioid numbers go down, and in doing so, I've personally attended to a large number of patients who've been directly harmed by involuntary tapers and discontinuations. . . [VA Clinician and Researcher] Health systems-level challenges. Poor interconnectivity between electronic medical records and other data monitoring systems results in gaps in data and creates burden for providers. Patient healthcare often occurs across multiple systems (e.g., military and civilian), with the result that data on opioid prescribing may be incomplete and difficult to access. Although state-based PDMPs gather pharmacy data on controlled prescriptions, experts reported that neither VA nor MHS health systems were consistently reporting to the PDMP, limiting their utility and requiring providers to run multiple queries. Barriers to PDMP use (e.g., complicated login) can increase the time required for each query of available prescribing data, while concerns about the quality of available data may reduce the perceived value of such queries. Chronic pain patients may enter a system through a variety of specialties, e.g., primary care, emergency medicine, or mental health. Panelists noted that this raises additional challenges for care coordination, follow-up, and identification of potential opioid misuse.
What I see is people are being referred to mental health, "Oh, it's addiction so we are going to refer them to mental health," and then mental health is like, "Oh, pain is related to opioids-that's not my thing, Access to high-quality chronic pain management is lacking for many patients. Panelists also reported that gaps in healthcare coverage and access impact the delivery of chronic pain care nationally, citing a lack of available specialty providers, particularly in behavioral health, lack of reimbursement available for non-pharmacological pain care, and lack of resources to support providers in "humanely and appropriately" weaning patients off opioids when needed.
[A] lot of people have drunk the Kool-Aid that we need to treat pain differently and there has been a big emphasis on alternatives to opioids, and in psychology, behavioral medicine, [as] an obvious component of that. The problem is that patients don't have great access to the treatments that we know are evidence-based, that we know work, that doctors would like to plug their patients into. This is the fundamental problem in my mind.

. .[Civilian and VA Clinician and Research]
Provider-level challenges. Provider time for clinical care is a scarce resource. Despite general agreement that delivering effective, patient-centered care for patients with chronic pain requires time and attention, panelists noted that a lack of time remains a core challenge at the provider level.
The issue that you're going to see regardless though is time is kind of the capital of the primary care clinic. . .. That conversation [about pain management] takes time, so it's really going to be a big cultural change to less spectator medicine and more proactive, you know, getting in there and fixing these problems, which can be very hard to do in certain clinics.

[Military Clinician]
Many providers are overloaded by workflow and information demands. Knowledge and technologies to support high-quality chronic pain care are available and proliferating; however, panelists noted that the utility of existing tools to support quality pain care is limited by providers' cognitive overload.
One of the issues we deal with, day in and day out, is all the "good idea" fairies that add to the primary care doc's load. . .I have to be aware of not just this good idea, but the good idea about what language they want to use and what their healthcare literacy score is and whether they're safe at home. Providers may face pressure from patients to provide opioid medications. Panelists also pointed out that patients with chronic pain may actively request an opioid prescription, and may be unsatisfied with a provider who refuses this request.
[I]f I'm seeing a patient who has already developed a strong idea that opioids are somehow the-either helpful or the token by which they judge whether their doctor is engaged with them, so, "You provide this to me or I'm gonna accuse you of being not committed to me." [VA Clinician and Researcher] In addition to having to cope with an unhappy patient, providers in some settings of care may be penalized if they receive low ratings on patient satisfaction related to a reluctance to prescribe opioids.
Patient satisfaction is directly tied to do you receive pain medication, opioid specifically, and do you receive antibiotics. . ..[I]n the civilian world that's directly tied to your ability to be paid to your maximum amount and to keep your job even.

[Military Clinician]
Patient-level challenges. Patients may have unrealistic expectations regarding chronic pain management. At the level of the patient, panelists pointed to the fact that many patients seeking care for chronic pain do so with the expectation that medication is the best treatment option, or that effective care will achieve immediate results or result in a life without pain.
[O]ne of the things that is the most difficult issue to deal with, with patients who come in with a complaint of pain is managing their expectations. Some people are just not going to get complete pain relief, and they are expecting to be perfect. [Civilian Clinician]

Recommendations
Recommendations outlined by panelists are described below (see also Table 3), beginning at the systems level and narrowing to those targeting providers and/or patients. Increase cross-system learning and collaboration. Panelists noted that the recent crisis within pain care has been accompanied by a surge in the available knowledge and resources for chronic pain management across VA, military, and civilian health systems. This growth provides an opportunity to engage in cross-system learning and collaboration: We [MHS] work very closely with the VA, and we've been drafting off a lot of their successes and I don't call them failures. Even times where the VA has had a rough go of it, they've led the way in a lot of the things that the [MHS] is benefitting by. . .they've shared a lot of experience with us.

[Military Policymaker]
Expand access to integrative chronic pain care and treatment for substance use disorders. In response to access gaps described above, panelists emphasized the importance of increasing availability of services for integrative chronic pain care and substance use disorder treatment.
[F]rom our perspective and experience, it's a matter of having a menu of options that are non-opioid, non-medication, available. So when you do choose opioids, it's appropriate and required, and not just because there's nothing else to offer the patient. Develop clinical decision support (CDS) for providers without increasing cognitive load. Panelists discussed at length what kinds of tools and resources would be of greatest utility for providers, generally preferring CDS that support effective practice in chronic pain and opioid risk assessment, treatment planning, and outcomes and safety monitoring. Panelists also offered the caveat that more information is not always better, particularly if it increases demand on providers' time and attention, and emphasized the importance of making tools simple and functional: . . .I would suggest giving people tools to help with [opioid management] and not overwhelming them. . .. very nice, simple, provider-focused tools that are meant for really busy providers to be able to do that. [VA Researcher and Policymaker] Train providers in safe and effective chronic pain care. Panelists noted that providers across an array of specialties provide care for acute and/or chronic pain, often with minimal As far as something actionable. . .some healthcare leader in that institution needs to be involved in this. This will need consistent reinforcement, otherwise you're going to get folks who do this great for about a month or two, and then they're like, "Ahh it's too much work, I'm done," and then they never see it again. training. Panelists pointed to need for provider training in application of clinical practice guidelines, use of existing information resources (e.g., PDMPs, CDS) and risk management practices (e.g., urine drug screens, opioid tapering), and effective communication with patients (e.g., realistic goal-setting, shared decision-making).
[G]ive the providers that are [asking for help] some education and tools in how to address this type of patients in the future. [Civilian Clinician] Implement new measures with appropriate training and support to achieve uptake and sustainment. Panelists also noted the importance of ensuring new tools (e.g., CDS), guidelines, and other resources (e.g., screening or referral technologies) are introduced alongside appropriate training and other implementation strategies, such as backing from leadership, clinical champions, and/or continuing education and feedback, in order to ensure uptake and sustainment.
[S]ome healthcare leader in that institution needs to be involved in this. This will need consistent reinforcement, otherwise you're going to get folks who do this great for about a month or two, and then they're like, "Ahh it's too much work, I'm done," and then they never see it again. So you have to have that sustainment. . .that continued implementation tail.

[Military Clinician]
Educate and engage patients in safe and effective chronic pain management. Panelists offered a vision for supporting patients to achieve reduced pain intensity and improved functioning and quality of life. They cited the need for resources to help patients set and achieve realistic goals for pain management.

Discussion
We conducted qualitative discussions with a panel of national experts to identify key challenges and recommendations in reducing risks associated with prescribed opioids. Panelists provided insight into challenges across multiple levels of the U.S. health system, including: the technical complexity of treating chronic pain; the fraught national climate around opioids; the need to integrate surveillance data across a fragmented health system; a lack of access to nonpharmacological options for chronic pain care; and the difficulties inherent in asking providers and patients to negotiate treatment for a complex condition in brief clinical encounters, often without adequate knowledge on either side of the risks, benefits and/or availability of chronic pain care options. Panelists' discussion of core challenges for providers underscore findings of other recent studies, in which providers described navigating a "tightrope" between the push to avoid opioid prescriptions for chronic pain and pressure from patients to prescribe [10,[24][25][26]. A study of patient satisfaction scores by Sites et al. [27] found that, among patients with musculoskeletal pain, those receiving opioid medications were more likely to report being highly satisfied with their care. Patients with chronic pain, meanwhile, may report feeling reliant on opioids even when expressing ambivalence regarding their benefit [28,29]. Given this, prescribers' sense of walking a tightrope becomes understandable, particularly given the tense national climate described by participating experts.
In the course of these expert discussions, panelists collaboratively imagined a healthcare delivery system that would ensure access to behavioral health and other non-pharmacological therapies for chronic pain, achieve interconnectivity and care coordination within and across health systems, utilize thoughtful CDS and tools to support providers in real-time assessment of likely treatment risks and benefits, and provide education and training for both providers and patients. They recommended that clinical teams would also facilitate patients' access to consultation and treatment for substance use disorders [30], while permitting opioid use as part of safe and appropriate pain management for those most likely to benefit. Establishment of the new Defense Health Agency may enhance opportunities to improve collaboration across the MHS, civilian sector, and VA.
These findings point to challenges and opportunities at multiple levels, beginning with providers and patients themselves. Repeated studies have shown deficits in patient-provider communication around opioid prescribing for acute and chronic pain [31][32][33], consistent with those described by panel members. These deficits persist despite availability of a variety of resources to educate providers to engage in "healthy dialogue" with patients around appropriate use of opioids [34]. Participating experts also highlighted continued need to identify and disseminate evidence-based practices for comprehensive pain management, to include best practices for opioid prescribing and screening and referral for substance use disorders, as needed. Although there is growing consensus regarding recommendations for chronic pain care [6], opioid risk mitigation strategies remain complex. Providers require ongoing training and support to implement recommendations. Practical tools for decision support in busy clinical settings [10,35] are required, but must be feasible for health care providers to use routinely amid juggling a continuum of other health issues. At the broader system level, panelists acknowledged that access to both nonpharmacologic pain care and substance use disorder treatment remains constrained in many-and particularly rural-areas. There are well-recognized disparities in the availability of specialty providers and evidence-based treatment options across the country [36][37][38][39][40]. In addition, many healthcare insurance plans limit coverage for nonpharmacologic chronic pain treatment options, despite efforts to increase parity in coverage of mental health and substance use care under the Affordable Care Act [41][42][43]. Healthcare coverage plans may also be inconsistent with CDC and other guidelines for pain care, which explicitly recommend multimodal pain care and integration of CBT. Although one survey of state Medicaid agencies found that most reported providing at least some coverage for non-opioid pain treatment [44], a cross-sectional study of 45 plans representing Medicaid, commercial, and Medicare Advantage insurers found that few covered acupuncture or psychological interventions like CBT for chronic non-cancer pain [45]. Evidence from the VA suggests, however, that patients are receptive to nonpharmacologic pain management when those options are made available [44]. Thus efforts to ensure healthcare coverage aligns with guidelines for effective pain care are likely to remain an important area for future work. Creative solutions must also be identified to address geographic impediments to comprehensive pain management.
The method of qualitative panel discussions used here has a number of strengths, which include reflecting expert knowledge and incorporating diverse national, cross-system, and multidisciplinary perspectives [46]. Given constraints on expert panelists' time, we did not adopt a formalized process of developing consensus (e.g., Delphi method); as a result, weighting recommendations in terms of their perceived importance or feasibility awaits further investigation. Our semi-structured discussion approach, however, allowed for more in-depth exploration of topics than is often possible with consensus-focused methods [47]. Additional limitations included modest sample size and incomplete representation across diverse medical specialties and health systems.
These findings point to significant challenges facing the U.S. health system as opioid overdose remains a leading cause of injury and death [48,49]. Leaders and policymakers will need to weigh the needs and challenges specific to their own agencies and institutions, and select appropriate next steps in light of existing initiatives, perceived needs, and available resources. VA, for example, has implemented several of the recommendations listed here as part of its Opioid Safety Initiative and other programs, and has seen a corresponding drop in high-dose opioid prescribing among Veterans [50]. As of late 2017, VA was required to report to state PDMPs, thereby increasing the accessibility of prescribing data across VA and non-VA systems [51]. Despite this progress, continued reductions in opioid-related morbidity and mortality are likely to require a public health perspective and broader collaboration across our nation's healthcare systems to identify gaps and "loopholes" in chronic pain care, and to develop targeted, patient-centered solutions for opioid risk mitigation [52].
Significant federal resources have been brought to bear on prevention, treatment and recovery from opioid use disorder. However, without addressing these and other issues, we risk failing to make the fundamental changes in our health care system that will prevent the next opioid 'crisis'.