Does PCR-based pathogen identification reduce mortality in bloodstream infections? Insights from a difference-in-difference analysis.

Abstract

Background: Bloodstream infections (BSI) are associated with high mortality rates, particularly when caused by resistant pathogens. Reducing the delay in diagnosis and initiation of appropriate treatment is crucial for improving clinical outcomes. The implementation of polymerase chain reaction (PCR) tests in the diagnostic process offers a promising approach to achieving quicker identification of pathogens, thereby potentially reducing mortality associated with BSI.

Methods: A difference-in-differences analysis was performed within a New York City hospital system, comparing mortality risk between patients with enterococcal BSI before and after the adoption of BCID2 PCR testing, using as control those with methicillin-sensitive S. aureus BSI, for which diagnostic protocol has been unchanged.

Results: The study included 548 inpatients; 164 diagnosed with vancomycin-resistant enterococci (VRE) BSI and 384 with MSSA BSI. The mean 30-day mortality risk difference in the period post-intervention estimated in our difference-in-differences model was -6.03 per 100 (95% CI: -10.35 to -1.7), with event study plots suggesting minimal deviation from parallel trends in the pre-treatment period.

Conclusions: Findings suggest that introduction of BCID2 PCR testing for enterococcal bloodstream infections (BSI) may be associated with a reduction in mortality, however, interpretation of the effects must be approached with caution given the relative imprecision of estimates. Further research with larger samples is essential to establish a definitive conclusion on the impact of rapid PCR testing on mortality in BSI. This is an innovative approach using causal methods to evaluate interventions aimed at the improvement of infection control and antimicrobial treatment strategies.