"We treat the patient, not the lab results:Root cause analysis of blood culture underutilization in neonatal care, Benin

Background: Blood culture is the gold standard for diagnosing neonatal infections, but its use is poorly documented in sub-Saharan Africa. This study assessed blood culture utilization and identified the root causes of its underutilization in a neonatal unit of a tertiary hospital in Benin. Methods: This study was part of an embedded case study and used the Iceberg model as its analytical framework. We reviewed the medical records of 385 consecutive neonates hospitalized for at least 48 hours between August 2024 and May 2025. These records were triangulated with 15 weeks of participant observation, 25 individual interviews and one group discussion with managers, clinicians and infection prevention and control staff. Results: Among the 211 neonates with suspected maternofoetal infection, 47 (22.3%) had a blood culture prescribed. Forty-two blood cultures were performed and 15 were positive. All 7 Gram-negative isolates were resistant to ampicillin, cefotaxime and ceftriaxone, while amikacin and carbapenems remained active. The main root cause was a mental model shared by clinicians: “we treat the patient, not the lab results”. This expression, reported by health workers, reflects a preference for clinical judgment over systematic laboratory confirmation. This mental model was reinforced by the absence of a prescription protocol, by a fixed-price admission kit that left the choice of tests to the prescriber, and by accountability gaps. Conclusion: Blood culture was critically underutilized in this neonatal unit. Underutilization will persist unless the mental models that drive health workers’ practices are mapped and addressed.