Objectives: The optimal duration of antibiotic therapy for uncomplicated Gram-negative bacteraemia remains undefined. Our objective was to compare clinical outcomes of receiving short (7-10 days) versus prolonged (>10 days) durations of antibiotic therapy for children with uncomplicated Gram-negative bacteraemia.
Methods: We conducted a retrospective cohort study of children with uncomplicated Gram-negative bacteraemia at The Johns Hopkins Hospital between 2002 and 2012. We estimated the risk of bacteraemic relapse among children who received short versus prolonged durations of antibiotic therapy using 1:1 nearest neighbour propensity score matching without replacement prior to performing regression analysis.
Results: There were 170 matched pairs that were well balanced on baseline covariates. The median duration of therapy in the short and prolonged courses was 10 days (IQR 10-10) and 14 days (IQR 14-17), respectively. The 30 day mortality was similar between the groups (OR 1.12; 95% CI 0.96-1.21). A prolonged duration of antibiotic therapy did not reduce the relapse risk compared with shorter durations (adjusted hazard ratio 0.67; 95% CI 0.35-1.27). Similarly, each additional day of antibiotic therapy was not protective against relapse risk (adjusted hazard ratio 0.99 per additional day; 95% CI 0.92-1.03). There was a trend towards an increased subsequent risk of candidaemia in children receiving longer treatment durations (hazard ratio 2.44; 95% CI 0.97-6.19).
Conclusions: Antibiotic treatment for more than 10 days for uncomplicated bacteraemia in children does not reduce the risk of microbiological relapse compared with shorter-course therapy, but may be associated with an increased risk of candidaemia. Our findings need to be confirmed in a larger, prospective study.
Keywords: candidaemia; duration; paediatrics; resistance.