# Effectiveness of an antimicrobial stewardship program using an automated antimicrobial surveillance system based on indication for antimicrobial administration

## TL;DR

A new automated system helps track antibiotic use by infection type, leading to reduced carbapenem and antipseudomonal antibiotic use in a hospital setting.

## Contribution

An automated surveillance system for antimicrobial stewardship that tracks usage by indication, leading to targeted interventions.

## Key findings

- Carbapenem use decreased by 0.69 per 100 patient days after the intervention.
- Antipseudomonal agent use also significantly decreased following the program.
- High carbapenem use in specific infections was identified despite fewer patients with hematological malignancies.

## Abstract

Conventional antimicrobial surveillance measures usage in days of therapy (DOT). However, assessments using DOT alone often ignore indication-specific details, which limits the identification of targets for antimicrobial stewardship (AS) interventions. Therefore, we have developed an automated surveillance system (the Antimicrobial and Patient Background Surveillance System [APBSS]) in order to analyze antimicrobial use based on indications, and evaluated the short-term effectiveness of an AS program using APBSS surveillance data, as an initial evaluation of the newly developed system.

Using an interrupted time series design, we calculated the carbapenem DOT per 100 patient days as the primary outcome by comparing the pre-intervention (October 2022 to September 2024) and post-intervention (October 2024 to September 2025) periods. During the intervention, the AS team presented the APBSS surveillance data, including indication-specific antimicrobial use, patient characteristics, annual trends at our hospital, and multicenter comparisons. The antipseudomonal agent DOT was a secondary outcome.

Patient characteristics did not differ substantially between the periods. APBSS surveillance data identified key issues, including high carbapenem use in specific infections (e.g. Gram-negative rod bacteremia), which occurred despite the lower proportion of patients with hematological malignancies compared with other hospitals. However, the intervention significantly reduced carbapenem DOT levels (level change: −0.69/100 patient days, p = 0.017). The antipseudomonal agent DOT also significantly decreased (level change: −1.7/100 patient days, p = 0.004).

The AS program that used APBSS surveillance data improved the use of carbapenems and antipseudomonal agents in the short-term. Thus, our initial evaluation has suggested that AS programs that use APBSS surveillance data can contribute to the appropriate use of antimicrobial agents.

The online version contains supplementary material available at 10.1186/s40780-025-00528-0.