Comparison of efficacy of a 7-day versus a 14-day course of intravenous antibiotics in the treatment of uncomplicated neonatal bacterial sepsis: study protocol of a randomized controlled non-inferiority trial

Abstract

Background

Neonatal sepsis is a global public health problem. There is no consensus regarding the optimum duration of antibiotics for culture-proven neonatal sepsis. Published randomized controlled trials (RCTs) comparing shorter versus longer courses of antibiotics provide low-quality evidence with serious risk of bias. We hypothesized that among neonates with uncomplicated culture-proven sepsis, antibiotic duration of 7 days is not inferior to 14 days.

Methods

This is a multi-centric, parallel-group, stratified, block-randomized, active-controlled, non-inferiority trial with outcome assessment blinded. Stratification is by center and birth weight. Neonates weighing ≥1000 g at birth, with blood-culture-proven sepsis (barring Staphylococcus aureus and fungi), without conditions warranting > 14 days antibiotics, and who clinically remit, are enrolled in the RCT on day 7 of administration of sensitive antibiotics. They are randomly allocated to no further antibiotics (intervention arm: total 7 days) or 7 more days of the same antibiotics (control arm: total 14 days). Allocation is concealed by opaque, sealed envelopes. The primary outcome is “definite or probable relapse” within 21 days after antibiotic completion. Secondary outcomes include definite and probable relapses at various timepoints until day 35 post-randomization, secondary infections, and adverse events. The neonatologist adjudicating probable relapses and lab personnel are blinded. Three hundred fifty subjects will be recruited in each arm, assuming a non-inferiority margin of 7%, one-sided alpha error 5%, and power of 90%. Analysis will be per protocol and by intention-to-treat. An independent Data Safety Monitoring Board monitors adverse events and will perform one interim analysis when 50% of expected primary outcomes have occurred or 50% of subjects have completed follow-up, whichever is earlier. O’Brien-Fleming criteria will be used to stop for mid-term benefit and Pocock’s to stop for mid-term harm. A priori subgroup analyses are planned by birth weight categories, gram-stain status of pathogens, and radiological pneumonia.

Discussion

This trial will provide evidence to guide practice regarding optimum duration of antibiotics for culture-proven neonatal bacterial sepsis. If a 7-day regime is proved to be non-inferior to a 14-day regime, it is likely to reduce hospital stay, costs, adverse effects of drugs, and nosocomial infections.

Trial registration

Clinical Trials Registry India CTRI/2017/09/009743. Registered on 13 September 2017.