555 - Pediatric Blood Culture Contamination: Risk Factors and Implications for Healthcare Utilization

Publication Number: 555.5744

Giselle Fernandes, Kaiser Permanente - Oakland Medical Center, Emeryville, CA, United States; Peter B. Cooch, Kaiser Permanente - Oakland Medical Center, Oakland, CA, United States; Kelsey Childress, Kaiser Foundation Hospital - Roseville Women's and Children's Services, Roseville, CA, United States; Nancy Fong, Kaiser Permanente Northern California, Oakland, CA, United States; Michael Hartmann, The Permanente Medical Group, Castro Valley, CA, United States; Marilyn Kwan, Kaiser Permanente Division of Research, Pleasanton, CA, United States

Background: Contaminated blood cultures, accounting for 25-50% of positives in the U.S., lead to unnecessary healthcare use, antibiotic exposure, and costs. While benchmarks target contamination rates under 1%, pediatric risk factors remain poorly understood, underscoring the need to identify these factors to improve outcomes and reduce avoidable expenses.

Objective: Within a pediatric population, we aimed to identify risk factors for contamination by age and setting and estimate the burden of unnecessary healthcare utilization due to contaminants, specifically emergency department (ED) visits.

Design/Methods: This multicenter, retrospective cohort study included pediatric patients with blood cultures drawn in inpatient, outpatient, or ED settings across a large healthcare system from 2014-2022. Blood cultures positive for common skin flora were considered contaminants after excluding patients with risk factors for true bloodstream infections. Multivariate analysis compared risk factors and healthcare use across positive, negative, and contaminated cultures.

Results: Of 89,948 blood cultures analyzed, 1,594 (1.8%) were contaminants. Infants aged 29-90 days had significantly higher odds of blood culture contamination vs school-aged children, with over a 5-fold increase (OR=5.66; 95% CI: 4.64-6.91). Age group and collector type interactions were also key risk factors. Across all age groups, phlebotomists had the lowest contamination rates, while ED nurses had 2.8x higher contamination odds (OR=2.79; 95% CI: 2.47-3.15). However, in infants 0-28 days, inpatient nurses had significantly lower odds of contamination than phlebotomists (OR=0.36; 95% CI: 0.28-0.46). Race and median household income had no significant impact on odds of contamination. In outpatients, 31.8% of patients with contaminated cultures had an ED callback within five days vs only 3.0% of those with negative cultures. Using an absolute risk reduction of 29.4 and an estimate of $3,400 per ED visit, contaminated cultures costs $979 each with $1,554,970 in excess costs over 10 years.

Conclusion(s): In this multicenter retrospective cohort study, age and collector type were strongly associated with blood culture contamination. Phlebotomist-drawn cultures reduced contaminants in most pediatric patient groups, though inpatient nurses achieved the lowest rates in neonates, who had the highest odds of contamination overall. Our findings highlight the economic impact of contaminants, as over 20% of outpatients with contaminated cultures were called back to the ED. This study emphasizes the need for systemic changes to reduce contamination and optimize resource use.

Figure 1.
Contamination Rates by Patient Age, Collector, and Clinical Location

Figure 2.
Contrast Odds Ratios for Contamination by Collector Stratified by Age Group

Figure 3.
ED Callbacks for Outpatients Stratified by Blood Culture Result