Blood culture diagnostic yield in a paediatric emergency department

Abstract

Objectives: Blood culture (BC) is one of the most commonly used microbiological tests in the paediatric emergency department (PED) despite its lack of immediate diagnostic utility. Our objectives were to uncover the reasons for BC in the PED, to review adherence to the current protocol for this technique, and to analyse the diagnostic yield (DY) of BC in localized bacterial infections.

Patients and methods: Retrospective study including patients younger than 18 years with a BC obtained in the PED of a tertiary care hospital in 2011. Patients with bacteraemia risk factors (immunosuppressant condition, long-term vascular catheter, prosthetic valve or ventriculoperitoneal shunt, and hospitalization or aggressive diagnostic or therapeutic procedure in the previous week) were excluded.

Results: Two thousand and sixty-two BCs were included. Fever without source (FWS) (34.3%) and pneumonia (10.1%) were the main indications for BC. Twenty-five per cent of BCs did not adhere to protocol: FWS (115 cases), mononucleosic syndrome (83 cases) and acute gastroenteritis (78 cases) represented half of these. The global DY by BC was 2.0%. Streptococcus pneumoniae was the main isolated microorganism (41.4%) and pneumonia was the main associated pathology (22%). All of the BCs with non-adhered-to protocol were negative. All of the BCs obtained for noncomplicated bacterial skin and soft tissue infection were negative.

Conclusion: FWS is the main reason for BC in the PED. One-quarter of the BCs were not indicated, with the DY null in these cases. Due to the low DY in certain localized bacterial infections, a strict individualized ordering of BC in these cases should be considered.