Optimizing Antibiotic Strategies With a Clinical Pathway Approach: Enhanced Initiation and De-escalation in Community-Acquired Pneumonia.

BACKGROUND

Community-acquired pneumonia (CAP) remains a leading cause of hospitalization, mortality, and health care expenditure worldwide. Inappropriate use of broad-spectrum antibiotics contributes to antimicrobial resistance, adverse drug events, and increased costs. Antibiotic stewardship and standardized clinical pathways are key strategies to improve prescribing practices.

METHODS

This quality improvement (QI) initiative was conducted at a 240-bed tertiary care teaching hospital from January 2023 through January 2025. A CAP-specific digital clinical pathway was developed using Curbside Health® and integrated into the electronic health record (EHR), accessible through a structured order set to support real-time clinical decision-making. The intervention included provider education and iterative refinement through 2 Plan-Do-Study-Act (PDSA) cycles. Community-acquired pneumonia cases were identified using ICD-10 coding and antibiotic order indication. Outcomes were analyzed at the system level using aggregated EHR data. Statistical analyses were performed using chi-squared, Fisher's exact, and t tests or Mann-Whitney U tests, as appropriate.

RESULTS

A total of 880 patients were included in the pre-intervention period and 1345 in the post-intervention period. Following implementation, broad-spectrum antibiotic use significantly decreased (vancomycin: 71.4% to 26.2%, P < .0001; piperacillin/tazobactam: 65.2% to 28.8%, P < .0001). Concurrently, guideline-concordant antibiotic use increased (ceftriaxone: 13.3% to 56.1%, P < .0001; azithromycin: 44.2% to 71.1%, P < .0001). Clinical pathway utilization increased from 4.0% to 73.5% (P < .0001). Estimated annual cost savings were approximately $650 000.

CONCLUSION

Integration of a CAP-specific digital clinical pathway into the EHR was associated with improved antibiotic stewardship and reduced health care costs. These findings support the feasibility of EHR-embedded decision support tools to optimize antibiotic prescribing. Patient-level clinical outcomes and demographic data were not assessed, consistent with the system-level design of this QI initiative, which warrants further evaluation in future studies.