How to assess early-onset neonatal sepsis? Comparison of three detection strategies

Early-onset neonatal sepsis (EONS) can cause significant morbidity and mortality, especially if it is not detected early. Given the decrease in its incidence in the past few decades, it is important to find a balance between reducing the use of diagnostic tests and continuing to detect affected patients. We compared 3 detection strategies in patients with risk factors (RFs) for infection: laboratory screening (S1), the Neonatal Sepsis Risk Calculator (S2) and clinical observation (S3).

Retrospective observational study in neonates born at 34 weeks of gestation or later and with RFs or symptoms compatible with EONS. We analysed outcomes in our unit with the use of laboratory screening (S1) and compared them with the other two strategies (S2 and S3) to contemplate whether to modify our protocol.

The study included 754 patients, and the most frequent RFs were prolonged rupture of membranes (35.5%) and maternal colonization by Streptococcus agalactiae (38.5%). Strategies S2 and S3 would decrease the performance of laboratory tests (S1, 56.8% of patients; S2, 9.9%; S3, 22.4%; P < 0.01), hospital admissions (S1, 11%; S2, 6.9%; S3, 7.9%; P < 0.01) and the use of antibiotherapy (S1, 8.6%; S2, 6.7%; S3, 6.4%; P < 0.01). Sepsis was diagnosed in 13 patients, and it would have been detected with S2 and S3 except in 1 patient who had asymptomatic bacteriemia by Enterococcus faecalis. No patient with mild and self-limited symptoms in whom antibiotherapy was not started received a diagnosis of sepsis later on.

Close clinical observation seems to be a safe option and could reduce the use of diagnostic tests, hospital admission and unnecessary antibiotherapy. The watchful waiting approach in patients with mild and self-limiting symptoms in the first hours post birth does not appear to be associated with failure to identify sepsis.