RAPID-CARE: Rapid Antibiotic Optimization in the ICU After Implementation of a Pneumonia Multiplex PCR Test—A Real-World Evaluation

TL;DR

A rapid PCR test for ICU pneumonia improved antibiotic decisions by quickly identifying pathogens and reducing unnecessary broad-spectrum use.

Contribution

Real-world evaluation of a pneumonia PCR test's impact on antibiotic stewardship in ICU patients with suspected infections.

Findings

The PCR test showed high sensitivity and negative predictive value for detecting pathogens in ICU patients.

The test directly influenced antibiotic decisions in over half of cases, especially when prior antibiotics were not used.

False positives were common in patients with prior antibiotic therapy targeting detected organisms.

Abstract

Background/Objectives: Lower respiratory tract infections (LRTIs) are frequent in the intensive care unit (ICU) and drive empiric broad-spectrum antibiotic use. Rapid multiplex PCR assays may improve pathogen detection and stewardship compared with conventional culture. We evaluated the real-world impact of the BioFire® FilmArray® Pneumonia Panel Plus (FA-PNEU®) on antimicrobial management in suspected nosocomial LRTI. Methods: This was a single-centre, prospective observational cohort study conducted in a tertiary ICU (Madrid, Spain) between April 2021 and March 2025. Adult patients with suspected hospital-acquired pneumonia (HAP), ventilator-associated pneumonia (VAP), or ventilator-associated tracheobronchitis (VAT) were included if paired respiratory samples underwent FA-PNEU® and conventional culture (CC). Diagnostic accuracy and prescribing changes were analysed. Results: A total of 344 samples from 236 patients were included. FA-PNEU® demonstrated high sensitivity (93.4%) and negative predictive value (97.9%) but moderate specificity (65.0%) and low positive predictive value (36.5%). False positives occurred in 85.8% of patients with prior antibiotic therapy targeting the detected organism. Antibiotic management was considered directly influenced by FA-PNEU® when any prescribing decision (initiation, escalation, de-escalation, or discontinuation) explicitly followed the panel’s results rather than other clinical or microbiological information. Using this definition, FA-PNEU® directly influenced antibiotic therapy in 57.6% of cases, while in 17.7%, prescribing was instead guided by a suspected alternative infection. In patients without prior antibiotics, treatment initiation or withholding was fully concordant with FA-PNEU® results, while in those already receiving therapy, 60.8% underwent modification, two-thirds in agreement with the panel. Conclusions: In critically ill patients with suspected nosocomial LRTI, FA-PNEU® provided rapid, high-sensitivity diagnostics that substantially influenced antimicrobial prescribing. Its greatest value lies in ruling out bacterial infection and guiding stewardship, though results must be interpreted within the full clinical and microbiological context.

Figures