Improving alcohol-related care in small-medium primary care practices: An evaluation of an adaptation of the SPARC trial intervention for small-medium sized practices

Abstract

Introduction

The Sustained Patient-centered Alcohol-Related Care (SPARC) trial demonstrated that 6 months of practice facilitation, decision support in electronic health records (EHRs) and performance feedback, increased identification and treatment of unhealthy alcohol use (UAU) in primary care (PC). The Michigan SPARC (MI-SPARC) study tested an adaptation of SPARC for small-medium PC practices.

Methods

PC practices were recruited for and participated in alcohol-related quality improvement (2/2020–2/2023). Outcomes collected for quality improvement were used for this evaluation, with data collected manually (“manual practices”) or electronically (“electronic practices”). The prevalence of EHR-documented brief intervention (BI) and AUD medication treatment (primary outcomes), and alcohol screening and AUD diagnosis (secondary outcomes) were measured at baseline and 6-months. Secondary data from formative evaluation and practice surveys were analyzed using PRISM domains: external environment, recipients, implementation infrastructure, and intervention.

Results

25 practices enrolled; 14 completed data collection. Neither primary outcome was consistently monitored or collected by practices. Mean prevalence of documented screening increased from 20 % to 55 % (manual practices) and from 3 % to 20 % (electronic practices). The mean prevalences of documented AUD diagnosis at baseline and follow-up, were 1.4 % and 3.8 % (manual) and 0.1 % and 0.05 % (electronic). At follow-up, 12 practices reported screening with validated questionnaires, and 13 and 8 offering BI and AUD medications respectively. Barriers identified were low resources, small PC teams, low EHR functionality, intervention complexity, stigma, and COVID-19.

Conclusion

Despite adaptions for smaller PC practices and improvements in screening, MI-SPARC did not increase documented BI or AUD medication treatment, largely reflecting mismatch between intervention complexity and implementation infrastructure in PC practices.