WIP 14 - Neonatal Sequential Organ Failure Assessment (nSOFA) Score in Predicting Survival in Neonatal Sepsis: A Systematic Review and Meta-Analysis

Publication Number: WIP 14.7705

EMAN HASHIM, University of Toronto Temerty Faculty of Medicine, TORONTO, ON, Canada; Nishkal Persad, University of Toronto Temerty Faculty of Medicine, Mississauga, ON, Canada; Bonny Jasani, The Hospital for Sick Children, Toronto, ON, Canada; Dany E. Weisz, University of Toronto Temerty Faculty of Medicine, Toronto, ON, Canada

Background: Sepsis is a major cause of death among neonates. Identifying those at risk of mortality may help prioritize intensive care support and ultimately improve survival. A major clinical gap is the absence of an accurate and universally accepted prediction model. The Neonatal Sequential Organ Failure Assessment (nSOFA) score comprises respiratory, cardiovascular and hematological indices to predict mortality among neonates. Although several studies have evaluated its ability to predict mortality in neonates with sepsis, there is variability in its diagnostic accuracy, leaving uncertainty about its applicability.

Objective: To perform a systematic review and meta-analysis of the nSOFA scores performance in predicting survival among neonates with sepsis.

Design/Methods: MEDLINE, EMBASE and the Cochrane library will be searched till November 30th 2024 for studies that evaluated the predictive performance of the nSOFA among hospitalized term and preterm neonates with sepsis. The primary outcome will be all-cause in-hospital mortality. Screening of studies, data extraction and risk of bias assessment will be independently performed by two authors. The Prediction model Risk of Bias Assessment (PROBAST) tool will be used to assess methodological quality. Random-effects meta-analyses, using restricted maximum likelihood estimation and the Hartung-Knapp-Sidik-Jonkman method, will be performed to estimate the summary discriminatory performance and calibration. Multivariate meta-analysis will be performed to jointly synthesize summary estimates of discrimination and calibration to increase precision. Heterogeneity will be evaluated using the tau2 and I2 statistics and explored using subgroup analyses (e.g. term vs. preterm; early vs. late onset sepsis). Meta-regression analyses will be performed using the transformed estimate of the model performance measure as the dependent variable and study level characteristics (e.g. mean gestational age, timing of nSOFA score application) as the explanatory variable. Sensitivity analyses will be performed excluding studies at high risk of bias.