Factors Associated with Prolonged Antibiotic Therapy in Neonates with Suspected Early-Onset Sepsis

Abstract

Early-onset sepsis (EOS) is a rare but profoundly serious bacterial infection. Neonates at risk of EOS are often treated with antibiotics. The start of empiric antibiotic therapy can successfully be reduced by the implementation of the EOS calculator. However, once started, antibiotic therapy is often continued despite a negative blood culture. To decrease the burden of antibiotic therapy, it is necessary to know whether the clinician’s reasons are based on objective factors. Therefore, we performed a retrospective single-centre cohort study to identify the factors associated with prolongation of antibiotic therapy in neonates with suspected EOS but a negative blood culture. Maternal, clinical, and laboratory data of neonates with a gestational age of ≥32 weeks, admitted between January 2019 and June 2021, were collected. Among neonates with a negative blood culture, we compared neonates with prolonged (≥3 days) to neonates with discontinued (<3 days) antibiotic therapy. The clinician’s reported reasons for prolonging therapy were explored. Blood cultures were positive in 4/146 (2.7%), negative in 131/146 (89.7%), and not obtained in 11/146 (7.5%) of the neonates. The incidence of EOS was 0.7 per 1000 neonates. Of the 131 neonates with a negative blood culture, 47 neonates (35.9%) received prolonged antibiotic therapy. In the prolonged group, the mean gestational age was higher (38.9 versus 36.8 weeks), and spontaneous preterm birth was less prevalent (21.3% versus 53.6%). Prolonged treatment was associated with late onset of respiratory distress, respiratory rate, hypoxia, apnoea and bradycardia, pale appearance, behavioural change, and elevated CRP levels. The most reported reasons were clinical appearance (38.3%), elevated CRP levels (36.2%), and skin colour (10.6%). Prolonging empiric antibiotic therapy despite a negative blood culture is common in suspected EOS. Clinical signs associated with prolongation are uncommon and the reported reasons for prolongation contain subjective assessments and arbitrary interpretations that are not supported by the guideline recommendations as arguments for prolonged therapy.