Abstract
IMPORTANCE: A national shortage of blood culture bottles affected approximately half of US hospitals, necessitating changes in blood culture practices. OBJECTIVE: To determine the association between restrictive blood culture stewardship measures and patient outcomes at a pediatric hospital. DESIGN, SETTING, AND PARTICIPANTS: A retrospective cohort study was performed to determine the association between new restrictive stewardship measures and blood culture practices in a quaternary care pediatric medical center from July 1, 2023, to January 31, 2025. All children with blood cultures collected were included. EXPOSURES: The following restrictive measures were enacted: (1) a 48-hour restriction on all aerobic blood cultures, (2) a 7-day restriction on all anaerobic blood cultures, (3) authorization to order an additional aerobic blood culture every 7 days, and (4) pooling of all central vein catheter lumens in 1 blood culture bottle. MAIN OUTCOMES AND MEASURES: The primary outcome was the monthly blood culture positivity rate, stratified and normalized per 100 emergency department (ED) visits and per 100 patient-days. RESULTS: During the study period, a total of 18 132 blood cultures from 5063 patient visits (median [IQR] age, 5.6 [1.1-12.4] years; 2744 [54.2%] male; 358 [7.15%] Black, 3013 [58.5%] Latino/a/x/Hispanic, and 720 [14.2%] White patients) were collected from August 1, 2023, to July 31, 2024 (preintervention period), and 6449 blood cultures from 2495 patient visits (1391 [55.8%] male, median [IQR] age; 5.5 [1.1-12.5] years; 191 [7.7%] Black, 1452 [58.2%] Latino/a/x/Hispanic, and 375 [15.0%] White patients) were collected from August 1, 2024, to January 31, 2025 (postintervention period). Restrictive stewardship measures were associated with a significant decrease in the blood culture collection rate per 100 ED visits (cumulative change, -24.1%; 95% CI, -38.4% to -8.9%; P = .01) and per 100 patient-days (cumulative change, -45.8%; 95% CI, -64.7% to -26.9%; P < .001). In the ED and inpatient units, the blood culture positivity rate remained stable, with a 14.4% (95% CI, -23.1% to 52.0%) and 27.8% (95% CI, -13.7% to 69.3%) cumulative increase, respectively. There was no significant change in hospital mortality due to septic shock (0.13% vs 0.15%) or mean length of stay (5.4 days vs 6.4 days), but there was a significant decrease in readmissions (3.3 vs 2.7; cumulative change, -27.5%; 95% CI, -7.6% to -47.6%; P = .02). CONCLUSIONS AND RELEVANCE: The implementation of restrictive blood culture measures, including pooling lumens and decreasing testing cadence, successfully reduced blood cultures collected without decreasing blood culture positivity rates and without an increase in mortality secondary to septic shock, readmission rates, or length of stay.