This paper describes the evaluation of the first known pain coach program implemented in the ED. Throughout the first year of implementation, adaptations were made to enhance the fit of the intervention and its implementation with the clinical context, external forces, and patient and clinician needs and preferences. Our findings are promising and support the need for further investigation of program effectiveness and implementation outcomes in the ED and other settings.
In the first 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 pain-related ICD-10 codes admitted during the implementation period; however, this is an imprecise comparison as diagnostic codes alone are not sufficient to identify patients in need of pain management support (32). Therefore, it is difficult to determine if program recipients were representative of the overall eligible population or if factors known to influence pain management, such as disparities in analgesic administration and influence 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clinicians was especially important as use of features such as automated data import or copy-and-paste can result in outdated or inaccurate EHR data (35).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 benefits 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 final 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-efficacy, 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 first 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 specific 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 first 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 findings 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 specific 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 first three months of program implementation and subsequently declined, which may be attributable to several factors. First, patient eligibility criteria were refined in the first 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 identification 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.