# 147. Clinical impact of the BioFire® FilmArray® Pneumonia (PN) Panel in the Veteran Affairs healthcare system using a difference-in-difference (DiD) design

## TL;DR

This study shows that using the BioFire PN Panel in VA hospitals helped improve early antibiotic treatment and de-escalation for ICU pneumonia patients.

## Contribution

The study provides real-world evidence of the clinical benefits of the BioFire PN Panel using a difference-in-differences design in a large healthcare system.

## Key findings

- BioFire implementation was linked to better early appropriate antibiotic therapy.
- Antimicrobial de-escalation improved with BioFire use in ICU pneumonia patients.
- No significant effect on 30-day mortality was observed.

## Abstract

The BioFire® FilmArray® PN Panel (BioFire) improves time to pathogen detection, but real-world evidence of clinical impact is limited. This study evaluated the clinical outcomes associated with BioFire implementation in ICU patients within the US Veterans Affairs (VA) system using a quasi-experimental design and difference-in-differences(DiD) analysis.

Comparison of Baseline Characteristics Between BioFire® and Control Groups

Unadjusted Differences in Outcomes between the BioFire and Control groups in the Pre-Post BioFire Implementation Periods Overall and Among Patients with a Positive Result for a BioFire PN Panel Bacterial Pathogen

A retrospective, multicenter, quasi-experimental study was conducted in the VA system (2016–2024). Hospitals implementing the BioFire were matched 1:1 to non-BioFire (Control) hospitals by time, complexity, and geographic region. Patients were first stratified by group (BioFire vs. Control) and implementation timeframe (Pre vs. Post) and were included if they had an ICU admission, radiographic evidence of PN, respiratory culture performed (and BioFire test in Post-BioFire group), survival ≥24 hours after index PN culture, and receipt of ≥1 antibiotic within 2 days of index PN culture. Outcomes included early antibiotic de-escalation (reduction in spectrum score < 24 h of index culture), early appropriate therapy (receipt of appropriate therapy < 24 h of index culture/Biofire test), and 30-day all-cause mortality. The effect of BioFire implementation was estimated using a DiD Poisson regression model, with and without inverse probability of treatment weighting (IPTW) adjustment overall and among patients with a positive result for a BioFire PN Panel bacterial pathogen on a BioFire PN Panel test or respiratory culture.Clinical Outcomes Associated with BioFire® Implementation in Difference-in-Difference Poisson Regression Analyses

Clinical Outcomes Associated with BioFire® Implementation in Difference-in-Difference Poisson Regression Analyses

A total of 11,746 patients across 36 facilities met study criteria. Many significant baseline differences were noted pre- vs. post groups, while Pre-BioFire-Pre-Control and Post-BioFire-Post-Control were similar (Table 1). Unadjusted outcome differences in outcomes between the BioFire and Control groups are shown in Table 2. In the DiD regression analyses (parallel trends assumption was met), BioFire implementation was associated with improvements in receipt of early appropriate therapy and de-escalation, but not 30-day mortality in overall analyses. Similar results were observed in patients with a positive BioFire/culture result (Table 3).

BioFire PN Panel implementation was associated with improved antimicrobial optimization in ICU patients with PN in the VA system.

Thomas Lodise, Jr., PharmD, PhD, GSK: Advisor/Consultant