Background
Blood culture (BC) testing is a critical tool in diagnosing bloodstream infections in pediatric patients. However, in low- and middle-income countries (LMICs), BC often shows limited diagnostic yield and carries a substantial risk of contamination, which challenges its cost-effectiveness and clinical utility. Identifying clinical and demographic factors that predict higher rates of true positive BC results could inform targeted diagnostic strategies, optimize resource allocation, and improve antimicrobial stewardship in pediatric hospital care.
Study Design and Methods
The prospective ACORN2 surveillance network collected standardized clinical and microbiological data from 19 hospitals across nine African and Asian countries, focusing on pediatric patients hospitalized with suspected infections. Eligible participants were children under 18 years who were prescribed parenteral antibiotics and had at least one BC sample taken during their hospital stay. Sociodemographic details, clinical severity signs, suspected infection syndromes, and BC results were recorded and analyzed.
The primary outcome was BC positivity with confirmed true pathogens, excluding contaminants and uncertain pathogens. Mixed-effects logistic regression models were employed, incorporating study site and patient identity as random effects to adjust for clustering and repeated measurements, to identify factors independently associated with positive BC results.
Key Findings
From 26,407 pediatric infection episodes recorded, 17,815 (67%) included a BC sample, with 15,384 episodes meeting analysis criteria. Among these, true pathogens were isolated in 689 episodes (4.5%), contaminants in 1,399 (9%), and uncertain pathogens in 143 (0.9%).
Multivariable analysis revealed several significant predictors of true positive BC:
– Age group: Children aged 29 days to 12 months had a 33% increased odds (OR 1.33, 95% CI 1.06–1.66), and those 5 to 18 years had a 62% increased odds (OR 1.62, 95% CI 1.30–2.01), compared to the 1 to 4 years group.
– Clinical severity: Each additional clinical severity sign increased the odds by 29% (OR 1.29, 95% CI 1.18–1.40).
– Infection acquisition: Hospital-acquired infections were associated with a threefold increase in odds of positive BC (OR 3.05, 95% CI 2.30–4.06) compared to community-acquired infections.
– Suspected infection syndrome: Compared with respiratory infections, sepsis (OR 2.09, 95% CI 1.67–2.61), gastrointestinal/abdominal (OR 2.36, 95% CI 1.78–3.13), skin and soft tissue/bone infections (OR 3.64, 95% CI 2.57–5.14), and genitourinary infections (OR 2.22, 95% CI 1.39–3.56) showed higher likelihood of positive BC.
These findings confirm the overall low BC yield in pediatric patients within LMIC hospital settings and highlight specific clinical subgroups with enhanced probability of pathogen detection.
Expert Commentary
The ACORN2 study provides valuable granular insights into the epidemiology of pediatric bloodstream infections across diverse, resource-limited settings. Its large, multicenter design strengthens generalizability across Africa and Asia. The identification of age-related differences aligns with known variations in immune development and pathogen exposure in children. The strong association between hospital-acquired infections and BC positivity underscores the importance of infection control and targeted surveillance in inpatient settings.
Interestingly, the analysis corroborates that respiratory infections—which are common in pediatrics—have comparatively lower BC positivity, suggesting that indiscriminate BC testing in this group may not be cost-effective. Conversely, the higher odds observed in sepsis and other syndromes support prioritizing BC in these conditions.
Limitations include potential variability in sampling protocols and microbiological capacities across sites, which could affect contamination rates and pathogen detection sensitivity. The study also does not address the impact of prior antibiotic exposure on BC yield. Additionally, while the model adjusts for severity signs, it is possible that some clinical indicators might correlate with BC positivity due to unmeasured confounders.
Conclusion
This comprehensive multicountry surveillance study confirms the limited overall yield of blood cultures in hospitalized children in LMICs but identifies distinct subgroups for whom BC testing is more diagnostically beneficial. Diagnostic stewardship programs can leverage these findings to optimize BC utilization by focusing efforts on infants, older children, patients with multiple severity signs, hospital-acquired infections, and specific clinical syndromes such as sepsis, skin and soft tissue infections, and gastrointestinal infections.
By tailoring BC sampling based on risk stratification, healthcare systems in resource-constrained settings can improve the efficiency of diagnostic testing, reduce contamination and costs, and subsequently enhance antimicrobial stewardship and patient outcomes.
Funding and Trial Registration
This study was conducted under the ACORN2 clinical surveillance network with data collection supported by participating institutions and funding bodies as reported in the original publication. The clinical trial registration number and funding sources were not specified in the provided summary.
References
Ardura-Garcia C, Hopkins J, Lee SJ, Waithira N, Painter C, Ling CL, Roberts T, et al. Factors associated with positive blood cultures in children in nine African and Asian countries: the ACORN2 surveillance network. BMJ Glob Health. 2025 Oct 20;10(10):e020448. doi: 10.1136/bmjgh-2025-020448. PMID: 41120197; PMCID: PMC12542579.
Additional literature on blood culture yield in children and diagnostic stewardship principles can be consulted for broader clinical context and guidelines.