A Pilot Study Evaluating the Impact of an Algorithm-Driven Protocol on Guideline-Concordant Antibiotic Prescribing in a Rural Primary Care Setting

TL;DR

This pilot study tested an algorithm-driven protocol to improve appropriate antibiotic use in a rural clinic, showing a potential decrease in inappropriate prescriptions.

Contribution

The study introduces an algorithm-driven protocol to enhance antibiotic stewardship in rural primary care settings.

Findings

The intervention was associated with a 12.6% decrease in antibiotic prescriptions discordant with clinical guidelines.

The odds ratio for inappropriate prescriptions was 0.55, suggesting a potential benefit of the algorithm-driven protocol.

The study highlights the potential of such protocols in resource-limited settings, though longer study periods are needed for confirmation.

Abstract

Approximately 2.8 million cases of bacterial antimicrobial resistance (AMR) infections result in over 35,000 deaths annually in the U.S. AMR is driven largely by inappropriate prescribing of antibiotics, especially in clinics serving rural communities or underserved populations. Antibiotic Stewardship Programs (ASPs) improve prescribing practices, but many rural clinics lack fully functional ASPs. This pilot study evaluated the impact of an algorithm-driven protocol on antibiotic prescribing in a rural primary care setting. We conducted a pre–post quasi-experimental study at a Federally Qualified Health Center (FQHC), focusing on upper respiratory infections, urinary tract infections, and sexually transmitted infections. Eligible patients were enrolled in the study during their primary care visits. The primary outcome was the frequency of guideline-concordant treatment, analyzed using descriptive statistics and Chi-square tests. Among 201 patients (101 pre-intervention, 100 post-intervention), the pre-intervention group consisted of 77% females and 47% African Americans, while the post-intervention group consisted of 72% females and 46% African Americans. The intervention was associated with a 12.6% decrease in the number of antibiotic prescriptions discordant with clinical guidelines (37.6% to 25%) from the pre- to post-intervention periods. This corresponded to an odds ratio of 0.55 (95% CI: 0.30–1.01, p = 0.054). Although not statistically significant at α = 0.05, this numerical decrease suggests potential benefits of algorithm-driven protocols in improving antibiotic stewardship in resource-limited settings. Longer study periods may further elucidate these benefits.