Advisor(s)

Russell Griffin

Committee Member(s)

Joshua Stripling
Katia Bruxvoort

Document Type

Thesis

Date of Award

6-1-2026

Degree Name

Master of Science in Public Health (MSPH)

School

School of Public Health

Department

Public Health

Abstract

Expected practice interventions are a type of antimicrobial stewardship intervention in which local guidelines regarding antimicrobial prescribing are developed and released at an institutional level to establish an institutional standard of care. An expected practice intervention to encourage use of oral and long-acting IV (PO/LAIV) antibiotics for serious infections was performed at the University of Alabama at Birmingham Hospitals in November of 2022. We conducted a quasi-experimental pre/post study using a difference in differences approach to evaluate the effect of these guidelines on a primary outcome of length of stay (LOS) and secondary outcomes of inpatient antibiotic days, 3-month readmission, PO/LAIV antibiotic use, microbiologic failure, and adverse drug events. We also assessed key factors associated with use of PO/LAIV therapy receipt using bivariate techniques and multivariable logistic regression. Among patients with eligible serious infections during the study time period (N=139), we found a significant increase in the proportion of encounters in which PO/LAIV therapy was prescribed after the intervention (42.0% vs 60.0%, p = 0.03) as well as a reduction in readmission (46.4% vs 25.7%, p = 0.01) but no significant difference in LOS, inpatient IV antibiotic days, microbiologic failure, or adverse antibiotic-related events. Fully adjusted interrupted time series (ITS) of the primary outcome of LOS demonstrated no significant step change before versus after the guidelines release (Annual Percent Change Rate Ratio [APC RR], 0.74; 95% CI, 0.48-1.15). The use of PO/LAIV therapy was significantly associated with younger age (49.3% vs 57.5%, p = 0.003), fewer Elixhauser comorbidities (median 3 vs 5, p = 0.002), non-white race (49.3% vs 28.9%, p = 0.01), and more microbiologic failure (11.3% vs 1.5%, p = 0.04). Multivariable logistic regression revealed an independent association between nonwhite race and receipt of PO/LAIV antibiotics (aOR 0.35; 95% CI 0.16-0.79). Overall, expected practice intervention implementation was associated with an increase in PO/LAIV therapy use and a reduction in readmission, reinforcing existing evidence on its role as a valuable emerging antimicrobial stewardship tool; however, this is tempered by the nonsignificant change in our primary outcome of LOS and secondary outcome of inpatient IV antibiotic days.

Keywords

Antimicrobial stewardship;Expected practice;Interupted time series;IV antibiotics

Download

Included in

Epidemiology Commons

Share

COinS