Clinical decision support reduced respiratory culture use by 16 percent across 15 US pediatric intensive care units (PICUs) without evidence of adverse patient outcomes, suggesting that more selective testing could help laboratories and clinicians reduce potentially unnecessary microbiology investigations.
Endotracheal aspirate cultures are commonly used to investigate suspected ventilator-associated infections in children with artificial airways. However, their diagnostic specificity is limited because bacteria detected in respiratory samples may represent colonization rather than infection. Positive cultures can therefore contribute to overdiagnosis and unnecessary antibiotic treatment.
The multicenter cohort study, reported in Critical Care Medicine, evaluated a diagnostic stewardship program implemented through the BrighT STAR Respiratory Quality Improvement Collaborative between 2019 and 2023. Participating PICUs introduced locally adapted clinical decision support, primarily checklists or algorithms, to standardize when endotracheal aspirate cultures should be collected and discourage testing in patients unlikely to have a ventilator-associated infection. The tools were initially delivered on paper or hospital intranets rather than embedded in electronic medical records.
Across more than 199,000 ventilator-days, the adjusted mean monthly culture rate fell from 7.80 to 6.55 cultures per 100 ventilator-days after implementation, a 16 percent reduction. Cultures repeated within 3 days also declined, from approximately 13 percent to 10 percent.
Importantly, reducing testing did not lead to detectable increases in bronchoalveolar lavage cultures, length of stay, readmissions, sepsis, septic shock, or duration of mechanical ventilation. Antibiotic use also remained broadly unchanged.
The findings reinforce the potential value of diagnostic stewardship to guide which samples are sent for testing. Respiratory cultures can generate clinically ambiguous results when testing is performed in patients with a low likelihood of infection. Standardizing indications for collection may therefore reduce low-value testing while improving the clinical relevance of results reaching the laboratory.
The intervention produced variable reductions between centers, however, indicating that local practice and implementation influence effectiveness. The authors also noted that patient-level data were unavailable and rare safety events could have been missed. Further research is needed to determine how best to integrate stewardship into routine workflows and assess its sustainability and cost impact.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
