Impact Over 3 Years of a Family Medicine-led Addiction Medicine Curriculum for Medical Students

Opioid overdose deaths have increased in the United States from 49,660 in 2019 to 71,238 in 2021. ¹ Medication for opioid use disorder (MOUD) is an effective approach for opioid use disorder (OUD) that reduces mortality and morbidity in patients with OUD. ^2-7

As of 2018, 57% of rural counties in the United States had no waivered providers of buprenorphine. ^{8, 9} Despite the recent removal of buprenorphine waiver requirements, barriers to treating OUD persist. Physicians not prescribing buprenorphine reported these barriers: colleague resistance, absent specialty backup, and low confidence managing OUDs. ¹⁰

Our inspiration stemmed from this unmet need and the scope-of-practice potential of family physicians. To improve access to addiction care, we applied an innovative education model rooted in the principles of Project ECHO (Extension for Community Healthcare Outcomes). ^{11, 12} ECHO’s design consists of a hub of specialists, typically from at-large academic institutions, who teach spokes of family physicians in surrounding underserved areas through videoconferences. ¹¹ In opioid education, ECHO models have been associated with increased buprenorphine prescriptions and retention in OUD treatment. ^12-21

ECHO is an established tool for expanding a family physician’s scope of practice, and educators are seeking to model ECHO in undergraduate medical education. ²² Our model moved the hub of expertise from an academic institution to a rural community, where two family physicians were content experts with experience implementing addiction medicine care within a family clinic. These experts taught ECHO seminars to medical students participating in two family medicine-focused longitudinal integrated clerkships (LICs): the Rural Physician Associate Program (RPAP) and the Metropolitan Physician Associate Program (MetroPAP). ^{23, 24} These LICs are 9-month family medicine experiences where third-year medical students are embedded in a community clinic.

We compared our ECHO curriculum to the addiction medicine exposure within traditional clerkships through the change in pre- and postcurriculum survey responses across three cohorts of students. The primary outcomes were student self-rating of skill level and attitudes surrounding addiction care.

Our prospective cohort study compared outcomes between the ECHO and traditional curriculums in third-year medical students. This project was submitted to the University of Minnesota’s Institutional Review Board and approval was deemed unnecessary.

Our study sample consisted of third-year medical students at the University of Minnesota across three academic years (2018-2021). The ECHO group included 131 RPAP and MetroPAP students who participated in 16 sessions of our videoconferencing addiction medicine curriculum delivered by community physician experts each academic year. One-hour interactive lessons consisting of 2/3 didactic lessons and 1/3 case discussions occurred over the lunch hour every other week (see Appendix A). The traditional curriculum group consisted of 583 third-year clerkship students not enrolled in RPAP or MetroPAP. This group received the standard third-year clerkship education covering addiction medicine, which consisted of two 90-minute didactic presentations and case-based online modules.

We evaluated program effectiveness through a 16-question survey regarding students’ self-assessment of their knowledge and attitudes before and after the program—a Kirkpatrick Level 1 type of evaluation. We based the knowledge self-assessment portion of this survey on the SCAN-ECHO questionnaire used by previous authors who studied ECHO in relation to addiction medicine training. ¹⁵ We based the survey of student attitudes on current barriers to practicing physicians when treating OUD. ¹⁰

We used two sample t tests to test for differences in the mean scores between the control group and the intervention group at each time point. These results are summarized in Tables 1 through 3, shown in the next section. In addition, we compared the change in scores for the intervention group to the change in scores for the control group using generalized estimating equations with a repeated subject statement to account for the paired nature of the data. We carried out all analyses using SAS 9.4 (SAS Institute), and a P value <.05 was considered statistically significant.

The pre- and postsurvey response rates were 98.4% for the ECHO group and 22.8% for the control group. We needed to exclude certain survey responses for reasons listed in Figure 1. We excluded responses from 2019-2020 because survey data did not include baseline identifiers. After exclusions, the final sample size included 56 ECHO students and 25 traditional clerkship students (ie, 112 and 50 survey responses, respectively). The ECHO and control groups shared similarities, including belonging to the same class in medical school and being assessed at the same time of year. Assessing for a difference in the change in each group negated the need for demographic matching. At the preintervention time point, the survey results revealed no statistically significant differences between the groups (Table 1).

Compared to the traditional clerkship, ECHO program participants reported significantly higher self-perceived ability in managing addiction concerns (mean score of 3.2 vs 2.2 on a 5-point Likert scale, P<.001; Table 2). Both groups reported strong beliefs that addiction medicine belongs in primary care, and these scores did not change over time (Table 3).

Our data showed that ECHO students reported greater confidence in addiction medicine training than traditional clerkship students. This finding demonstrates that our family physician-led curriculum was superior to a traditional curriculum in increasing self-perceived ability in third-year medical students to deliver addiction care. These results are consistent with prior ECHO model studies, which showed increased confidence in treating OUD. ^13-16 In addition, the results demonstrated that attitudes among all students trended toward a belief that primary care has a duty to treat patients suffering from addiction.

The methodological weaknesses of the study included fewer responses among the control group and a lack of participant identifiers in year 2, which forced exclusion of that group from statistical analysis. In addition, we assessed self-perception of ability rather than an objective measure of competence, such as an exam.

Like ours, past ECHO projects have used specialist training to expand clinical care to areas of high need. The success of this curriculum is related to its longitudinal nature, use of real cases, and role-modeling of family medicine experts who successfully practice in addiction care. In future studies, adding a competency-based exam would allow for more objective outcomes data.