Sepsis prediction during outbreaks at neonatal intensive care units through body surface screening for Gram-negative bacteria: systematic review and meta-analysis

Abstract Objective This systematic review focusses on the prognostic accuracy of neonatal body surface screening during outbreaks caused by Gram-negative bacteria for prediction of sepsis. In a previous systematic review we reported that only limited evidence of very low quality exists regarding the predictive value of this screening under routine conditions. We aimed to investigate whether this is different in outbreak settings. Results We identified five studies performed during outbreaks in three countries, comprising a total of 316 infants. All studies were at high risk of bias. In outbreak settings, pooled sensitivity of body surface screening to predict sepsis was 98% (95 CI 60 to 100%), while pooled specificity was 26% (95% CI 0.5 to 96%). Evidence quality was low for all outcomes. Extending a previously published systematic review, we show here that in contrast to routine settings sensitivity of body surface screening for sepsis prediction is very high, while specificity is still insufficient. Surface screening appears to be a useful component of bundles of interventions used during outbreaks, but the evidence base is still limited. PROSPERO Registration Number: CRD42016036664.