Implementation of a novel emergency department pain coach educator program: First year experience and evaluation

Background: The ongoing opioid epidemic and rising number of patients with chronic pain have highlighted the need for alternative and integrative pain management approaches. A number of evidence-based nonpharmacologic pain management strategies are available; however, these approaches remain underutilized due to barriers such as time limitations, cost, and lack of clinician training. The aim of this work was to implement a nonpharmacologic pain coach educator program that addresses these barriers. We report an evaluation of the first year of program implementation in the emergency department of a large safety-net hospital. Methods: We implemented a multimodal pain coach educator program that included education on pain neuroscience and over-the-counter analgesic options, demonstration of integrative techniques, and nonpharmacologic toolkits for home use. Implementation strategies included electronic health record tools, training and promotion, clinical champions, and clinician recognition. We used the RE-AIM framework to guide evaluation of the first year of program implementation using data from the electronic health record, quantitative and qualitative program records, and patient-reported outcomes. Results: In the first year of program implementation 550 pain coach educator sessions were conducted. Upon immediate session completion, 61% of patients felt the program was helpful, 39% were unsure at the time, and none reported session was not helpful. Clinician feedback was overwhelmingly positive. Program cost per patient was $344.35. Adaptations to first year intervention and implementation strategies included modifications of session delivery timing for accommodation of clinical workflows, additions to program content to align with patient characteristics, and changes to patient identification strategies in response to the coronavirus 19 pandemic. Conclusions: The PAMI pain coach educator program provides a model for nonpharmacologic pain management programs which can be scaled up and adapted for other settings. This work demonstrates the importance of intervention and implementation strategy adaptations to enhance program reach and effectiveness.

particularly in the complex and fast-paced environment of the emergency department.
We found a novel pain coach program was successfully implemented in an emergency department with ongoing adaptations to enhance reach, effectiveness, and feasibility.
Weekly discussion of program challenges with a multidisciplinary team and timely modi cations in response contributed to the success of the program.
This work contributes to the literature by providing a model for implementation of pain education programs in emergency departments that can be adapted for other settings.

Page 3/25
Background Chronic pain affects more Americans than diabetes, cancer, and heart disease combined and results in an estimated $600 billion in annual costs related to healthcare, disability, and lost productivity (1). Efforts to address the burden of pain in the United States (US), combined with the introduction of new opioid formulations misrepresented as having low potential for abuse, led to widespread misuse of both prescription and nonprescription opioids by the early 2000s (2). This ongoing opioid epidemic in the US has resulted in over 500,000 deaths and cost the economy an estimated $1 trillion in 2017 alone (3,4).
Updated prescribing guidelines along with legislative and regulatory action have corresponded with reductions in opioid prescribing (5,6); however, patients are increasingly turning to illicit drugs, and opioid-related overdose deaths continue to increase (7,8). The coronavirus 19  pandemic has contributed to the rise in opioid-related overdoses and deaths in recent years. Increased inactivity, worsening mental health, increased substance use, and limited access to non-prescription pain management (e.g., physical therapy, medical procedures) can exacerbate pain conditions and lead to a return to opioid use (9).
There is a growing focus in the US healthcare system on delivering nonpharmacologic pain management interventions and reducing opioid use. A number of evidence-based nonpharmacologic approaches to pain management are available, including non-opioid medications, restorative therapies (e.g., physical or occupational therapy), behavioral approaches, and complementary and integrative health (e.g., massage, acupuncture, yoga, tai-chi) (10). Unfortunately, these approaches are often under-utilized in clinical practice or not covered by insurance and many patients lack knowledge of or access to these treatments (10)(11)(12)(13)(14).
Safe and effective pain management is particularly challenging in emergency departments (EDs). Pain is the most common reason for ED visits (15,16). While prescribing rates are declining, about 15% of ED patients receive opioid prescriptions (16,17). Barriers to effective pain management in ED settings include lack of healthcare professional pain education, lack of access to pain specialists and consistent primary care, time and work ow limitations, insu cient insurance coverage, inadequate patient knowledge about medication safety, and frustration and anger over "shu ing" of care with no improvement in pain (13,(18)(19)(20)(21)(22)(23)(24).
The Pain Assessment & Management Initiative (PAMI) was established in 2014 and aims to advance multimodal, safe pain management in healthcare systems to improve outcomes and reduce opioid risk (25). One component of PAMI is a new model pain coaching program, which incorporates patient education and a patient pain toolkit for use after discharge. This program is the rst known ED pain coach education program in the US. This paper reports the experiences and evaluation of the rst year of program implementation in the ED of a large safety-net hospital.

Page 4/25
The pain coach educator program provides patient education on nonpharmacologic, integrative, and overthe-counter analgesic options for pain management, along with a patient nonpharmacologic pain toolkit for use after discharge. Since initial program launch, there has been expansion to inpatient and other settings within the health system. This paper reports on experiences in the pre-implementation period (September -December 2020) and rst year of implementation (January -December 2021) in the ED setting. The project is registered with the University of Florida Quality Improvement Project Registry.

Setting
The PAMI pain coach educator program was implemented in the ED of a large (700 bed), urban, safetynet, not-for-pro t hospital with approximately 62,000 in-patient admissions annually, serving populations in Northeast Florida and South Georgia. The ED serves over 70,000 patients annually and is the region's only Level I Adult and Pediatric Trauma Center. Most ED patients identify as Black (62%), are insured by Medicare or Medicaid (54%) or are self-pay or charity city funded pay (24%).

Program Description
The pain coach educator program is intended for patients age 14 years and older with acute or chronic pain. Patients with psychosis, suicidal or violent behavior, incarceration, severe uncontrolled pain prior to medication, restrained or immobilized, or critically ill were excluded from the program. Patients were referred to the program by physicians or advanced practice providers through an Electronic Health Record (EHR) paging system, phone call, or verbal request from other healthcare professionals (e.g., nurse, physical therapist, pharmacist). Program staff also monitored the EHR tracking board to identify eligible ED patients. The pain coach educator then reviewed the EHR to assess the patient's relevant medical history to determine appropriateness for the program. When possible, the pain coach educator conferred with a member of the patient's healthcare team prior to and following the pain coaching session.
Pain coach educator program components are described in detail in a publicly available toolkit on the PAMI website (26). Brie y, sessions consisted of 1) patient education on basic pain neuroscience and prevention of acute to chronic pain transitions, 2) demonstration of integrative pain management techniques, 3) a review of options to improve pain and quality of life, and 4) provision of nonpharmacologic toolkit items and educational brochures, and 5) a review of appropriate OTC and topical analgesic pain management options. The program was intended to be delivered in a single session; however, it was possible for patients to participate in the program during a later admission, ED visit, or via telephone upon patient request. Coaching sessions were tailored based on an initial assessment, type of pain, and patient characteristics and preferences. Patients were provided with a variety of toolkit item options and educational brochures. Brochures covered 17 topics including OTC oral and topical medications, sleep, diet, and back exercises. Toolkit items included aromatherapy inhalers, stress ball symbolizing a pain management analogy (27), hot/cold gel pack, virtual reality cardboard viewer with suggested free virtual reality apps, wearable acupressure device, pain journal, and a QR code to pain management videos on the PAMI website (25). The pain coach educator customized toolkits for each patient based on their pain assessment, contraindications, patient interest, and patient characteristics (e.g., smart phone access, comorbidities). If the pain coach educator was unavailable, clinical team members could provide patients with toolkit items by accessing a stocked cart located within the ED clinical areas.

Pain Coach Educator Quali cations and Training
The position description for the pain coach educator speci ed minimum quali cations of a bachelor's degree in an appropriate discipline and four years of relevant experience or equivalent. Preferred quali cations included a master's degree, clinical experience, and professional experience in patient education, pain management, neurobiology, or integrative medicine. In the rst year, the program was primarily delivered by a master's-level pain coach educator with formal training in integrative medicine and experience as an emergency medical responder. A small proportion of sessions (approximately 5%) were delivered by other team members with varied backgrounds (e.g., public health, health education) when the primary pain coach educator was unavailable. Orientation and training of the pain coach educators included 1) training in pain neuroscience education and pain management treatment options; 2) shadowing various pain-related healthcare professionals in the hospital system to understand the organizational structure, work environment, and clinical roles; 3) EHR training; and 4) review of current literature, textbooks, and online learning modules related to integrative and nonpharmacologic pain management.

Implementation Strategies
Four primary implementation strategies were used to promote uptake of the program in the ED during the rst year: 1) EHR system modi cations, 2) ongoing training and promotion activities, 3) clinical champions, and 4) clinician recognition. To facilitate referrals, the team worked with the health system EHR team to develop an "auto-page" function, which allowed providers to place EHR referrals tied to a speci c patient record. An EHR note template and owsheet were also developed for the program. PAMI staff delivered 22 clinician training and promotion sessions to 397 clinicians and staff during the rst year of implementation. Staff also promoted the program through a continuous on-site presence, rounding in clinical areas, email announcements, and visual cues (e.g., bulletin board, yers). The team engaged clinical champions in various roles (e.g., nurses, residents, physician assistants, physical therapists, pharmacists) to promote the program, serve as a liaison between the program and clinical staff, and provide feedback and updates on program initiatives. Finally, each month the top program utilizer was recognized on a message board in a high-tra c area of the ED, along with supervisor noti cation and a certi cate in their employee le. Further details on the program implementation and lessons learned are available on the PAMI website (25).

Evaluation
The Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) framework was used to guide evaluation. We chose RE-AIM because it is a well-established evaluation framework with the ability to capture patient, clinician, and setting-level outcomes (28). Selected outcome measures organized by RE-AIM are presented in Table 1. Data sources included the EHR, program records maintained by staff, and patient surveys.

Data collection
The pain coach educator documented delivery of each session in the EHR using a note template created speci cally for the program. The template includes the number of past-year ED visits for the patient, history of opioid use or new opioid prescription, toolkit items given, pain management topics discussed, brochures provided, and any additional session details documented narratively by the pain coach educator. The patient's last documented pain assessment was auto-populated into the note. The PAMI team also developed a structured data collection form including patient demographics, pain characteristics, opioid risk assessment, pain coaching and education session components (topics coached, toolkit items/brochures provided), patient feedback, challenges experienced during the session, and patient disposition including referrals given at discharge. The pain coach educator completed the structured data collection form following each visit. Separate project coordinators conducted chart reviews to complete the remaining sections of the form upon the patient's discharge from the ED and performed overall data veri cation. All data were stored and managed in REDCap.
Beginning in November 2021, a subsample of patients also completed a follow-up telephone survey administered by project coordinators one month after the coaching visit to assess at-home use of the coaching topics and toolkit items and gather general feedback. To identify patients for the follow-up survey, program staff generated a random sequence of patients from the prior month and contacted patients by phone one time until 10% completed the survey.
To evaluate clinician training sessions, a pre/post assessment measuring knowledge of nonpharmacologic toolkit items and nonpharmacologic pain management modalities was administered (see Appendix A). The pain coach educator and program staff also recorded informal feedback on the program obtained through discussions with ED clinicians and staff familiar with the program. Feedback was not audio recorded but staff attempted to record comments verbatim when possible. Program challenges and adaptations were systematically documented in weekly team meetings and described in monthly presentations to the funding agency. These presentations and meeting notes were compiled and reviewed by the program team to produce a nal list of challenges and adaptations organized by the Framework for Modi cation and Adaptations (FRAME) (29

Data Analysis
Descriptive statistics were generated for all quantitative measures. We calculated the proportion of referred patients who received the intervention and compared the characteristics of program recipients to all patients admitted to the ED with a pain-related ICD-10 codes (e.g., headache, migraine, musculoskeletal pain, low back pain, renal colic pain) during the evaluation period. Informal feedback from clinicians was reviewed and summarized; no formal qualitative analysis was conducted because responses were limited to brief statements. Cost per participant was calculated by dividing rst-year total expenditures by number of participants. Costs for the pre-implementation period were calculated separately.

Reach
Over the rst year of the program, 550 sessions were conducted with 549 unique patients. Characteristics of pain coach educator program recipients and all patients admitted with pain-related ICD-10 codes are presented in Table 2 Tables 3 and 4, respectively. Immediately after the session, most (61.1%) recipients reported the session was helpful, 38.9% were unsure, and no patients responded the session was not helpful. When asked which part of the session they found most helpful, patients most frequently identi ed both the coaching topics and the toolkit items as the most helpful components. Aromatherapy and the virtual reality viewer were the most commonly reported helpful toolkit items. At four-week follow up, all patients surveyed (n = 11) had used the materials and/or education received and gave high ratings (4 or 5 on a 1-5 scale) for the helpfulness of the program in managing pain at home. The majority (n = 8, 72.7%) of patients used program education and/or toolkit items daily.  Adoption Clinician comments about the program were overwhelmingly positive. Clinicians described reductions in pain-related anxiety and improved mood in patients receiving coaching sessions. One ED physician stated, "Everyone that talks to the pain coach seems much happier and uplifted." An ED pharmacist noted the pain coach educator "has demonstrated a positive in uence on the patient's level of empowerment in managing their pain. In conversing with patients, they express appreciation for [the pain coach educator] educating them on techniques such as mindfulness, distraction, virtual reality, and breathing which reduced their pain-related anxiety." The pharmacist also noted, "the pain coach is being well received by staff; [they] have done a good job of raising awareness of pain coaching services among staff." An ED nurse also described positive reactions to the program, but noted it was di cult to get program information to night shift staff due to limited program availability at night.

Implementation Fidelity
Session details are presented in Table 5. All patients were offered education on at least one pain management topic, which the patient could accept or decline. Patients were coached on the majority of educational topics offered, with rates of educational topics offered but not coached (i.e., declined by patient, insu cient time) ranging from 0.4% (virtual reality) to 10.8% (mind/body techniques). Most toolkits distributed included a stress ball, hot/cold pack, aromatherapy, and the PAMI postcard. No challenges were noted for 332 (60.4%) of sessions. The most frequent challenges related to the acuity of the patient's medical condition (e.g., nausea/vomiting, lethargy) and pain level (e.g., in too much pain to participate, cognition impaired by medication). Challenges and adaptations Table 6 presents program challenges experienced and resulting adaptations to the intervention and implementation strategies. Most modi cations occurred during the rst quarter of the implementation period.

Maintenance
The total number of sessions and those resulting from referrals by month are presented in Fig. 1

Discussion
This paper describes the evaluation of the rst known pain coach program implemented in the ED. Throughout the rst year of implementation, adaptations were made to enhance the t of the intervention and its implementation with the clinical context, external forces, and patient and clinician needs and preferences. Our ndings are promising and support the need for further investigation of program effectiveness and implementation outcomes in the ED and other settings.
In the rst year of implementation, 550 ED patients participated in the pain coach educator program. We found program recipients were more likely to be female and black compared to all ED patients with painrelated ICD-10 codes admitted during the implementation period; however, this is an imprecise comparison as diagnostic codes alone are not su cient to identify patients in need of pain management support (32). Therefore, it is di cult to determine if program recipients were representative of the overall eligible population or if factors known to in uence pain management, such as disparities in analgesic administration and in uence of culture and gender on pain reporting, played a role (33,34). In practice, program eligibility assessment relied on clinician consultation and review of both structured and narrative EHR data. In our experience, conferring with clinicians was especially important as use of features such as automated data import or copy-and-paste can result in outdated or inaccurate EHR data (35).
About 60% of patients reported the pain coach educator program was helpful at the conclusion of the session. As the intent of the intervention is to provide education and tools for patients to continue to use after discharge, it is unsurprising that bene ts may not be realized immediately for some patients.
Further, it is important to note that the remaining patients responded they were unsure of program helpfulness rather than asserting the program was not helpful. While follow-up assessments were newly introduced in the nal months of the evaluation period, we found promising results in the small sample that had completed the follow-up questions during year one. All patients had applied the education they received or used toolkit items after discharge and over 70% used the education or items daily. Although our evaluation of effectiveness was limited to assessment of patient satisfaction, previous research shows similar interventions are effective in improving pain intensity, pain interference, pain management self-e cacy, depression, and anxiety (36-38). Notably, one of these studies found a single session program to be as effective as an eight-session cognitive behavioral therapy program (38). In light of this previous work, additional research is needed to determine if the PAMI pain coach educator program produces similar results. In the second year of the program, procedures were implemented to attempt follow-up phone surveys on all patients receiving coaching sessions to better assess program effectiveness. Preliminary data from these assessments show high rates of continued utilization of skills 30 days after the session.
A valuable lesson learned during the rst year of implementation was the importance of timing when approaching patients. Patients have high expectations for timely and effective pain management, especially in the ED (39,40). Shortly after program rollout, the pain coach educator recognized that approaching patients prior to analgesic (oral or intravenous) administration was problematic in two ways. First, patients were unreceptive to the program because of their pain and frustration with lack of immediate pain relief. Second, some patients viewed the program as interfering with or replacing analgesic pain management. Modifying session timing to occur after analgesic administration combined with clear communication that participation in the program would not affect receipt of medications allowed for the delivery of more productive sessions to more patients.
Medical education programs typically devote minimal time to pain management and often focus on opioids (41,42). The PAMI program attempts to address this gap by providing training on nonpharmacologic pain management strategies. Surprisingly, we found that about a quarter of ED residents attending training sessions scored lower on post-test compared to pre-test assessments. Most incorrect answers on post-test assessments were in relation to two topics (belly breathing, hot/cold packs), indicating the need for improved education in these speci c areas. Subsequent training sessions focused on clarifying explanation of these topics.
We found the PAMI pain coach educator program cost $344.35 per patient during the implementation period (i.e., excluding start-up costs). Some personnel also supported program expansion to the inpatient setting during the rst year of ED implementation, so this cost is likely overestimated. Nonetheless, this cost is relatively low considering the potential for the program to reduce health care utilization and associated expenditures (43). Personnel costs were higher than originally anticipated due to the need for ancillary staff for program management, toolkit inventory, and administrative tasks. Some of this need stemmed from ongoing monitoring and evaluation activities, which could be scaled as appropriate for the implementation setting and resources (e.g., capacity to obtain evaluation metrics from the EHR). These initial ndings can be used to inform decisions to implement this and similar programs.
Nonpharmacological and integrative pain management approaches have been shown to be cost-effective in treating various types of pain; however, more cost-effectiveness studies speci c to pain management in ED settings are needed (44,45).
The number of monthly pain coach educator referrals and sessions delivered steadily increased in the rst three months of program implementation and subsequently declined, which may be attributable to several factors. First, patient eligibility criteria were re ned in the rst few months of program rollout in response to receipt of referrals for patients who were inappropriate for the program. Second, ED volumes at our institution historically decline in summer and increase in fall and winter. Finally, in late April 2021 the pain coach educator program expanded to include inpatient settings, which reduced program staff availability in the ED. The emergence of the COVID-19 Delta variant corresponded with decreased referrals; however, due to program staff efforts to identify eligible patients, the total sessions per month remained steady during this time. Given the substantial demands on clinician time and attention, particularly in ED settings (46, 47), proactive patient identi cation by program staff is recommended to supplement direct clinician referral methods when possible.
Strengths of this evaluation include the innovative nature of the pain coach educator program, use of implementation science planning and evaluation frameworks, and systematic documentation of program adaptations. Despite these strengths our approach has some limitations. First, we did not systematically document the number of patients ineligible or uninterested in the program. These procedures have subsequently been implemented in ongoing program activities. Second, clinician feedback was obtained informally rather than through rigorous qualitative methods. Finally, our economic evaluation was restricted to a budgetary impact analysis and did not include effectiveness outcomes.

Conclusions
This work describes a model for the design and implementation of nonpharmacologic pain management in the ED which can be scaled and adapted for other settings. Our ndings reinforce the importance of performing ongoing evaluation and adaptation of interventions and implementation strategies. Future work will evaluate the impact of the program on key outcomes such as admissions and costeffectiveness in the ED and other clinical settings. Availability of data and materials: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.