Optimizing Antibiotic Stewardship in a Regional Hospital Network

Thesis Abstract

Optimizing antibiotic stewardship within a regional hospital network to reduce inappropriate antibiotic use, curb resistance development, and improve patient outcomes is addressed in this study. The problem arises from fragmented stewardship practices across network facilities, limited diagnostic support, and variable prescriber adherence to guidelines, leading to elevated rates of empiric broad-spectrum antibiotic exposure and suboptimal treatment durations. The aim is to design and evaluate a coordinated stewardship intervention that harmonizes policies, enhances diagnostic stewardship, and strengthens prescriber engagement across the regional network. Specific objectives include (1) quantify baseline antibiotic utilization patterns and resistance trends across five tertiary- and secondary-care hospitals; (2) identify organizational, behavioral, and clinical determinants of stewardship gaps using a mixed-methods approach; (3) develop and implement a multilevel stewardship intervention incorporating guideline standardization, diagnostic pathway optimization, and real-time feedback; (4) evaluate the intervention’s impact on antibiotic consumption, appropriateness of prescribing, and clinical outcomes over 12 months; and (5) model the economic and epidemiological implications of sustained stewardship improvements. Methodologically, the study adopts a convergent parallel mixed-methods design conducted in a regional hospital network comprising five facilities serving diverse patient populations. The population includes all inpatients treated with systemic antibacterial therapy over a 12-month baseline period and a purposive sample of prescribers, infection control professionals, pharmacists, and microbiology staff (n=120 participants for surveys; 30 in-depth interviews). Data collection instruments combine (i) a retrospective audit of antibiotic prescriptions, indications, duration, de-escalation practices, and microbiology results; (ii) structured surveys capturing knowledge, attitudes, and practices related to antibiotic use, guideline familiarity, and perceived barriers; (iii) semi-structured interviews exploring contextual factors influencing stewardship behaviors; and (iv) observational checklists for adherence to diagnostic pathways. Validity and reliability are ensured through pilot testing, triangulation of sources, and inter-rater reliability checks for chart reviews (Cohen’s kappa ?0.80). Theoretical framing integrates the Health Belief Model to understand prescriber behavior, and the Consolidated Framework for Implementation Research (CFIR) to structure organizational determinants and implementation processes, complemented by the Theory of Planned Behavior for intention-to-prescribe. Data analysis employs (a) descriptive and inferential statistics to compare pre- and post-intervention metrics (paired t-tests, chi-square tests, and repeated-measures ANOVA where appropriate); (b) Interrupted time Series analysis to assess trajectory changes in antibiotic consumption and defined daily doses (DDD) per 1,000 patient-days; (c) multilevel mixed-effects regression models to account for clustering at hospital and ward levels; (d) thematic analysis of qualitative data using an inductive approach with NVivo 12, ensuring credibility through member checking; and (e) economic evaluation via cost-effectiveness analysis from the hospital perspective, calculating incremental cost per infection averted and per quality-adjusted life-year gained where feasible. Key expected findings include a measurable reduction in inappropriate empiric broad-spectrum use, shorter mean durations of therapy for common infections, improved rates of timely de-escalation, and enhanced concordance with guideline-concordant prescriptions. Improvements are anticipated in clinical outcomes such as reduced length of stay, lower incidence of Clostridioides difficile infection, and stabilized or reduced resistance rates for targeted pathogens. The study will elucidate critical organizational barriers and enablers, including leadership engagement, availability of rapid diagnostics, decision-support tools, and interprofessional collaboration, with CFIR mapping revealing specific domains most predictive of implementation success. The contribution to knowledge lies in providing a robust, transferable model for regional network-based antibiotic stewardship that combines standardization with tailored implementation strategies, reinforced by rigorous evaluation of clinical and economic outcomes. The main conclusion is that a coordinated, data-informed, multilevel stewardship intervention can produce durable improvements in antibiotic prescribing quality and patient outcomes across a regional hospital network while delivering favorable economic value. Recommendations emphasize sustaining real-time feedback mechanisms, integrating diagnostic stewardship with electronic health record prompts, ongoing education tailored to local resistance patterns, periodic audit-and-feedback cycles, and scalability considerations for other regional health systems with similar resource constraints.