An educational programme in neonatal intensive care units (SEPREVEN): a stepped-wedge, cluster-randomised controlled trial.

BACKGROUND:Patients in neonatal intensive care units (NICUs) are at high risk of adverse events. The effects of medical and paramedical education programmes to reduce these have not yet been assessed. METHODS:In this multicentre, stepped-wedge, cluster-randomised controlled trial done in France, we randomly assigned 12 NICUs to three clusters of four units. Eligible neonates were inpatients in a participating unit for at least 2 days, with a postmenstrual age of 42 weeks or less on admission. Each cluster followed a 4-month multifaceted programme including education about root-cause analysis and care bundles. The primary outcome was the rate of adverse events per 1000 patient-days, measured with a retrospective trigger-tool based chart review masked to allocation of randomly selected files. Analyses used mixed-effects Poisson modelling that adjusted for time. This trial is registered with ClinicalTrials.gov, NCT02598609. FINDINGS:Between Nov 23, 2015, and Nov 2, 2017, event rates were analysed for 3454 patients of these 12 NICUs for 65 830 patient-days. The event rate per 1000 patient-days reduced significantly from the control to the intervention period (33·9 vs 22·6; incidence rate ratio 0·67; 95% CI 0·50-0·88; p=0·0048). INTERPRETATION:A multiprofessional safety-promoting programme in NICUs reduced the rate of adverse events and severe and preventable adverse events in highly vulnerable patients. This programme could significantly improve care offered to critically ill neonates. FUNDING:Solidarity and Health Ministry, France.