Positive Blood Culture ≠ Bacteremia: New Decision Rule

Written by Joshua Belfer

Spoon Feed 
A simple 4-factor clinical decision rule accurately distinguished true bacteremia from contaminants in children with positive blood cultures, achieving 99% sensitivity while potentially reducing unnecessary hospitalizations.

Bacteremia or contaminant? A rule to guide the call back
False-positive blood cultures are a familiar and costly problem in the pediatric emergency room. Up to half of positive cultures represent contamination. This single-center retrospective study derived and validated a highly sensitive clinical decision rule to help clinicians differentiate true bacteremia from contaminants among children with positive blood cultures obtained in the ED.

Using two cohorts from a Canadian pediatric ED, authors analyzed 747 children with positive blood cultures, 51% of which were determined to have a contaminant. A Classification and Regression Tree (CART) model approach identified four key variables to stratify risk:

1. Gram stain showing gram-negative bacteria or gram-positive bacteria in pairs or chains
2. Time to culture positivity: <17 hours
3. Presence of an internal device (i.e. central line, prosthetic device)
4. Clinical suspicion of osteoarticular infection    

Children were classified as high (positive gram stain), moderate (any of the other 3 variables), or low (no variables) risk. In the validation cohort, the rule demonstrated 99% sensitivity (95%CI 94–100) and 60% specificity (95%CI 50–70), with only 1 of 83 bacteremia cases in the validation cohort misclassified as low risk. Among children discharged at the index visit, applying the rule would have led to a reduced hospitalization rate and a reduced antibiotic initiation rate, leading to a number needed to treat of 4.

This was a single-center, retrospective study, with both derivation and validation performed at the same institution, limiting generalizability of the findings prior to external validation. 

How will this change my practice?
Callbacks for positive blood cultures on discharged patients can send you into a tailspin. Should I have even sent the culture? Am I making this patient come back in the middle of the night for something not even real? This study offers the potential for some structure and guidance in that stressful moment. Even more appealing is the simplicity and generalizability of it. We know all four data points when the lab calls with that positive culture. While there is still a need for external validation and real-world implementation studies, this study provides an evidence-based framework to support decisions about who really needs to come back, be admitted, and receive antibiotics—and who likely doesn’t. I look forward to additional studies that hopefully will allow us to reduce unnecessary admissions, antibiotic exposure, and family disruption, while maintaining safety.

Source
Derivation and Validation of a Clinical Rule to Detect Bacteremia Versus Contaminants in Positive Pediatric Blood Cultures: A Retrospective Cohort Study. Ann Emerg Med. 2025 Dec;86(6):576-585. doi: 10.1016/j.annemergmed.2025.06.620. Epub 2025 Aug 14. PMID: 40810705.