1628 - Treatment Heterogeneity in Pseudomonas aeruginosa Pneumonia

Background: Serious bacterial infections present a unique challenge for studies of real-world evidence. Often, the causative organism is unknown during the initial period of treatment and clinical symptoms change day-to-day, which lead to multiple changes in therapy. While it is assumed approaches to treating specific infectious diseases are mostly similar, we’ve previously identified substantial treatment heterogeneity, even among organism-specific and site-specific infections.

Methods: Our retrospective cohort study included inpatients with positive P. aeruginosa from sputum and bronchoalveolar lavage cultures collected during VA medical center and community living center stays from 01/15-04/18. We included the first positive culture during the admission per patient. Daily antibiotic exposures were mapped from 3 days prior to the culture collection date until discharge or 30 days for longer stays. Heterogeneity was defined as patterns of antibiotic treatment (drug and duration) not shared by any other patient.

Results: Our study included 5,435 patients and 87.4% of patients had different patterns of antibiotic drug and duration. Among patients with changes in therapy (84.0%), 96.9% had different antibiotic treatment patterns, with a median time to first change of 1 day and median of 3 changes. When restricting the analysis to antibiotic classes (rather than drug), Gram-negative antibiotics, and anti-pseudomonal antibiotic classes, heterogeneity was 81.8%, 52.0%, and 48.7%, with median time to first change of 1, 3, and 3 days, and a median of 3, 2, and 2 changes, respectively.

Conclusion: Among inpatients with positive P. aeruginosa respiratory cultures, substantial heterogeneity was observed in the national VA Healthcare System. Even at the class level, and restricting the analysis to anti-pseudomonal antibiotic classes, approximately 50% of patients had different treatment patterns during their inpatient stay. Current methods to assess treatment do not adequately account for the extensive heterogeneity observed in infectious diseases and it remains unclear how local or national treatment guidelines affect heterogeneity.