# A five-year quasi-experimental study to evaluate the impact of empiric antibiotic order sets on antibiotic use metrics among hospitalized adult patients

## TL;DR

This study found that using antibiotic order sets in electronic medical records reduced antibiotic use and Clostridioides difficile infections in hospitalized adults over five years.

## Contribution

The study provides long-term evidence of the effectiveness of antibiotic order sets in improving antibiotic stewardship in a real-world hospital setting.

## Key findings

- Standardized antimicrobial administration ratio (SAAR) decreased significantly from 1.0 to 0.90 over five years.
- Antibiotic days of therapy per 1,000 patient days dropped from 4,884 to 3,939.
- Clostridioides difficile infection cases decreased significantly from 7.8 to 2.4 per 1,000 patient days.

## Abstract

Evaluation of adult antibiotic order sets (AOSs) on antibiotic stewardship metrics has been limited. The primary outcome was to evaluate the standardized antimicrobial administration ratio (SAAR). Secondary outcomes included antibiotic days of therapy (DOT) per 1,000 patient days (PD); selected antibiotic use; AOS utilization; Clostridioides difficile infection (CDI) cases; and clinicians’ perceptions of the AOS via a survey following the final study phase.

This 5-year, single-center, quasi-experimental study comprised 5 phases from 2017 to 2022 over 10-month periods between August 1 and May 31.

The study was conducted in a 752-bed tertiary care, academic medical center.

Our institution implemented AOSs in the electronic medical record (EMR) for common infections among hospitalized adults.

For the primary outcome, a statistically significant decreases in SAAR were detected from phase 1 to phase 5 (1.0 vs 0.90; P < .001). A statistically significant decreases were detected in DOT per 1,000 PD (4,884 vs 3,939; P = .001), fluoroquinolone orders (407 vs 175; P < .001), carbapenem orders (147 vs 106; P = .024), and clindamycin orders (113 vs 73; P = .01). No statistically significant change in mean vancomycin orders was detected (991 vs 902; P = .221). A statistically significant decrease in CDI cases was also detected (7.8, vs 2.4; P = .002) but may have been attributable to changes in CDI case diagnosis. Clinicians indicated that the AOSs were easy to use overall and that they helped them select the appropriate antibiotics.

Implementing AOS into the EMR was associated with a statistically significant reduction in SAAR, antibiotic DOT per 1,000 PD, selected antibiotic orders, and CDI cases.