Optimizing Blood Culture Draws Through Use of an Algorithm Can Reduce Utilization in a Community ICU

Abstract

Objective

To evaluate the implementation of a blood culture algorithm in a mixed medical-surgical ICU at a community hospital, and examine the association with blood culture utilization rate and patient outcomes. Design: A quasi-experimental study examining pre- and post-implementation periods. Setting: A 22-bed mixed medical-surgical ICU at a community hospital. Patients: Adult ICU patients were admitted between February 2022 and October 2024, excluding those with neutropenia (<500 cells/μL) or solid organ transplants. Intervention: Introduction of a multidisciplinary-developed blood culture algorithm designed to standardize ordering practices for new clinical events and clearance of bacteremia. Measurements and Main Results: Primary outcomes included blood culture event rates. Secondary outcomes were antibiotic days of therapy, mortality, and readmissions. Interrupted time series analysis using Poisson regression models were used to examine associations between the intervention and clinical outcomes. The intervention reduced blood culture event rates by 39% (IRR 0.61, 95% 0.49, 0.75) without significantly decreasing adverse events such as 90-day death incidence (5.7% vs 7.2%, p-value 0.44) and 30-day hospital readmission (11.0% vs 8.0%, p-value 0.11). Inappropriate blood culture rates also decreased. Conclusions: Implementation of a blood culture algorithm in a community ICU setting was associated with reduced blood culture utilization without compromising patient safety. The intervention may substantially reduce unnecessary blood cultures, addressing a key gap in diagnostic stewardship in non-academic settings.