Impact of an Educational Intervention on the Opioid Knowledge and Prescribing Behaviors of Resident Physicians

Objectives: The opioid epidemic is a multifactorial issue, which includes pain mismanagement. Resident physician education is essential in addressing this issue. We aimed to analyze the effects of an educational intervention on the knowledge and potential prescribing habits of emergency medicine (EM), general surgery (GS), and internal medicine residents (IM). Methods: Resident physicians were provided with educational materials and were given pre-tests and post-tests to complete. Descriptive statistics were used to analyze pre-test and post-test responses. Chi-squared analysis was used to identify changes between the pre-tests and post-tests. A p < 0.05 value was considered statistically significant. Results: Following the educational intervention, we observed improvement in correct prescribing habits for acute migraine management among emergency medicine residents (from 14.8% to 38.5%). Among general surgery residents, there was a significant improvement in adherence to narcotic amounts determined by recent studies for sleeve gastrectomy (p= 0.01) and laparoscopic cholecystectomy (p= 0.002). Additionally, we observed a decrease in the number of residents who would use opioids as a first-line treatment for migraines, arthritic joint pain, and nephrolithiasis. Discussion: Resident physicians have an essential role in combating the opioid epidemic. There was a significant improvement in various aspects of opioid-related pain management among emergency medicine, internal medicine, and general surgery residents following the educational interventions. We recommend that medical school and residency programs consider including opioid-related pain management in their curricula.


Introduction
The United States (US) opioid epidemic is a multifactorial crisis, with prescription opioids identified as a key contributor to opioid misuse and overdose deaths [1]. At the medical provider level, harm reduction techniques have focused on preventing an excess of prescription opioids from circulating in the community. Approaches to this have included legislative limits on prescription amounts for certain patient populations and state-mandated use of prescription drug monitoring programs (PDMPs) to regulate opioid dispensing [2][3]. While recent studies have provided specialty-and procedure-specific opioid prescribing recommendations based on patient consumption patterns and pain relief requirements, no formal prescribing guidelines exist to eradicate the provider uncertainty that stems from the fear of undermanaging patient pain [4][5][6][7]. Additionally, medical school and residency program curricula dedicated to key opioid and pain management topics are underwhelming in the context of the severity of the opioid epidemic. This has been attributed partly to a limited pool of faculty who feel qualified to teach these concepts and to a lack of standardized competencies driving curricular design [7,8].
Accordingly, residents across all medical disciplines are often underprepared to prescribe opioids for patient pain or respond to various opioid-related patient management scenarios [9][10][11]. The magnitude of this deficit is well-exemplified in a recent study surveying surgical residents at a large academic institution: 90% reported no formal training in best practices of pain management or opioid prescribing, despite reliance on opioids for postoperative pain management [11]. In response to this insufficiency in medical trainee preparation, residency programs have begun to incorporate opioid and pain management material into their curricula. Programs have used various educational models, and some have quantified the effectiveness of these didactics through methods such as survey data collection [11][12][13]. However, these interventions are typically implemented in a specialty-specific cohort, which limits group knowledge comparisons and the potential to evaluate standardized intervention effectiveness across a variety of medical disciplines. Raheemullah et al. conducted an opioid education intervention using pre-tests and post-tests among internal medicine residents and found improvement in knowledge and prescribing habits [14].
The purpose of this study was to investigate the impact of an educational presentation on resident knowledge and attitudes related to opioid prescribing and pain management, in internal medicine (IM), general surgery (GS), and emergency medicine (EM) residents. By implementing a standardized intervention designed to educate trainees on key concepts such as opioid crisis statistics, opioid prescribing laws, opioidrelated complications, and evidence-based opioid prescribing guidelines, we aimed to measure the success of this intervention at content delivery while simultaneously collecting data on the opioid and pain management education of our residents. The goal was to compare the effectiveness of this educational model at improving resident opioid and pain management knowledge, attitudes, and behaviors across several specialties and assess the feasibility of a generalized institutional approach to resident opioid education.
This article was previously posted on Research Square as a Preprint on August 17, 2021.

Materials And Methods
This study was determined to be exempt from institutional review board review by the Review Board of Thomas Jefferson University. A total of 46 IM, 17 GS, and 27 EM residents from all postgraduate years (PGYs) at Thomas Jefferson University Hospital in Philadelphia, Pennsylvania, were recruited by email to voluntarily participate in this study. The intervention was designed as a seven-minute pre-recorded lecture with accompanying pre-tests and post-tests. Tests were intended to assess resident opioid and pain management knowledge, attitudes, and behaviors at baseline and upon presentation completion. Lecture content consisted of opioid crisis statistics, opioid prescribing laws, opioid-related complications, and evidence-based opioid prescribing guidelines with practice recommendations modified for each specialty. The pre-tests and post-tests were designed by a team of physicians and medical students. Each test was identical for each group of residents, with differences only in case vignette content and prescribing guidelines between the three cohorts to provide residents with recommendations and scenarios relevant to their specific fields (see Appendix). The post-tests were taken shortly after the pre-tests. The complete pretests and post-tests for each specialty are available in the Appendix. The data from the pre-tests and posttests were organized and we performed descriptive statistics to quantify the responses. We used a Chisquared analysis to identify resident changes between baseline and completion of the educational intervention using IBM SPSS Statistics for Windows, Version 26.0 (Released 2019; IBM Corp, Armonk, New York). Additionally, we generated comparisons of performance measures across the three cohorts to identify specialty-specific trends. A p < 0.05 value was considered statistically significant.

Results
A total of 90 residents completed the pre-tests; there were 27 residents from EM, 17 from GS, and 46 from IM. There were 46 post-test responses from 13 EM residents, 13 GS, and 20 IM residents. The response rates between pre-test and post-test for EM, GS, and IM were 48%, 76%, and 43% respectively. The demographics for the residents are listed in Table 1.

Pre-test
Post-test Year  EM residents reported receiving education about opioids from various avenues and stages of training, including personal reading (10 residents (37%)), medical school (16 residents (59.3%)), and residency (22 residents (81.5%)). Regarding training previously received, four residents were very satisfied (14.8%), eight were satisfied (29.6%), 11 were neutral (40.7%), three were unsatisfied (11.1%), and one was very unsatisfied (3.7%). The EM resident prescribing habits and opioid knowledge are listed in Table 2. After receiving the educational intervention, the attitudes of EM residents to the statement "If I suspect someone is abusing opioids, I do not prescribe opioids to them" significantly changed (p=0.04).   Table 3.

Correct responses (%) Incorrect responses (%)
For a patient being discharged home after a sleeve gastrectomy, I would typically prescribe:  were neutral, three (6.5%) were satisfied, and two were very satisfied (4.3%). Following the study training, nine residents were unsatisfied (45%), nine were neutral (45%), and two were satisfied (10%). Following the educational intervention, there was an improvement in responses to multiple treatment scenarios, though none of this reached significance. These included treatment of acute episodic migraines according to American Headache Society 2015 Guidelines (45.7% to 70% prescribing naratriptan, p=0.11), improvement in prescribing habits for joint pain in a patient with a history of osteoarthritis (23.9% to 45%, p=0.14), and an increase in non-opioid management of nephrolithiasis in a patient with no history of GI bleed (62.2% to 70%, p=0.59) ( Table 4).
If I suspect someone is abusing opioids, I do not prescribe opioids to them.   Regarding the level of satisfaction with prior opioid training, there was a significant difference between specialties (p<0.0001). Almost half of all IM residents felt unsatisfied with their prior opioid training (unsatisfied or very unsatisfied -47.9%). Comparatively, 5.9% of GS residents and 14.8% of EM residents felt unsatisfied or very unsatisfied with their training. There was also a significant difference across specialties in the initial management of mild pain (p=0.005) and moderate pain (p<0.001). For moderate pain, GS residents (35.5%) were more likely to prescribe opioids than their colleagues in IM (2.2%) and EM (0%).

Discussion
The opioid epidemic in the US has progressively worsened. There are several historical factors that contributed to the rise of the opioid epidemic, including the classification of pain management as a human right [15] and a fifth vital sign [16], pharmaceutical marketing [17], and postoperative pain mismanagement [18]. To combat the epidemic, there has been growth in non-opioid treatments in pain management, such as nerve blocks, non-steroidal anti-inflammatory drugs (NSAIDs), and ketamine [19]. This study aimed to assess the baseline responses and the effect of a brief educational intervention on the knowledge and attitudes of EM, GS, and IM residents at a single institution.
This study found many significant opportunities for improvement in resident opioid education. Regarding previous opioid knowledge training, 45.7% of IM residents were unsatisfied with the quality of training they had received. This suggests an area of potential collaboration between residents and hospital administration to better equip trainees with the practical information and skills they need to safely and effectively manage pain.
With this brief intervention, there was an improvement of prescribing habits across all specialties. In EM, we observed a greater percentage of residents indicating knowledge that, per Pennsylvania state guidelines, seven days is the maximum duration of opioids that should be prescribed to an adult patient presenting to the ED with acute pain (22.2% to 30.8%). Additionally, there was an increase in the correct use of naratriptan for acute migraine management in the ED (from 14.8% to 38.5%). This particular scenario represents a key opportunity to reduce opioid use in exchange for a more efficacious medication. A study conducted by Colman et al. found that more than half of all patients presenting with migraines were treated with opioids as first-line therapy across four different hospitals [20]. Focusing on common clinical presentations like this, where treatment algorithms may be ambiguous for many providers, could greatly reduce the unnecessary use of opioids. Additionally, this effort is not meant to create a divide between providers and patients. Patients who have a history of drug misuse should receive the appropriate pharmacotherapy and psychosocial counseling to equip them with the tools to make effective change [21].
From the provider perspective, it is imperative to keep the patient's best interest in mind when treating someone struggling with drug dependence, without letting biases affect your judgement. We observed a change in perspective among EM residents. Initially, majority of residents would not prescribe opioids to someone who appeared to be misusing drugs (70.3%). After our intervention, the number of residents who agreed with this statement decreased to 30.8% (p=0.04). This change in perspective highlights the multifaceted and individualized approach needed for each patient, considering the dangers of both over and under-prescribing. Walter et al. observed significant improvement in knowledge and management of opioid use disorder among EM residents following an educational intervention [22].
In our study, we observed significant improvement for GS in prescribing habits, better conforming to narcotic amounts determined by recent papers, following common procedures such as sleeve gastrectomy (p= 0.01), and laparoscopic cholecystectomy (p= 0.002) [23,24]. A similar study conducted by Hill et al. found that an educational intervention effectively decreased the number of opioids prescribed to patients following general surgery procedures [25].
Among IM residents, there was a decrease in participants who wanted to use opioids as a first-line treatment for migraines, arthritic joint pain, and nephrolithiasis. While these findings may not reach statistical significance, the increased percentage of correct responses indicates improvement of knowledge. The recommended first-line treatment for acute migraine includes NSAIDs and triptans. Opioid use in migraine treatment has not shown to have significant improvement so they are not recommended as initial treatment [26]; however, studies such as Bigal et al. have found that opioids were commonly used in clinical practice for migraine treatment (20.8%) [27]. A possible solution in this gap between recommendations and clinical practice can be educational interventions such as this study to target specific clinical situations that are confusing for providers or commonly treated inappropriately with opioids when good alternatives exist.
Potential limitations in this study can be attributed to the study design. Since our study focused on survey responses, the data largely depended on completion of both pre-tests and post-tests. There was a discrepancy in response rates between the two tests, likely due to survey fatigue and the demands of residents' schedules. Additionally, this study took place during the coronavirus disease 2019 (COVID- 19) pandemic, which placed considerable stress on resident physicians [28]. In order to boost survey responses, we sent reminders via email, had participating residents from each department make announcements at weekly meetings, and sent other team members to attend departmental conferences. Despite our best efforts, however, we were unable to improve these response rates. Additionally, our data is from survey answers and not real-world clinical actions. Given residents are largely constrained in their medication prescribing practices by the desires and preferences of supervising attendings, we did not feel studying their prescribing behaviors would yield meaningful results. Resident physicians are an integral component in battling the opioid epidemic. With these findings, we encourage medical schools and residency programs to integrate training on the effective use of non-opioid pain treatments into their curricula.

Conclusions
The opioid epidemic is a multifaceted issue that can be attributed to many causes. Resident physicians are a key resource in combating the opioid epidemic. We observed significant improvement in opioid knowledge and prescribing habits among all residents following the specialty-specific educational interventions. Therefore, we recommend that medical school and residency programs consider integrating opioid-related pain management strategies throughout their curricula.