The risk prediction models of neonatal necrotizing enterocolitis in China: a systematic review

Abstract

AbstractObjectiveSystematically evaluate the risk prediction model for neonatal necrotizing enterocolitis (NEC) in China, providing reference for clinical work and future research.MethodsWe searched Chinese and English databases were systematically searched to focus on NEC risk prediction modeling studies.The search time ranged from database establishment to 25 August 2023.Two researchers independently screened the literature and extracted information.Then risk of bias and applicability were assessed by using the Prediction Model Risk of Bias Assessment Tool.ResultsA total of 10 papers involving 12 NEC risk prediction models were included, which is focusing on the populations of preterm infants mostly, building the methods of models diversity, predicting factors discrepancy widely.conclusionThe existing NEC risk prediction models in China have good predictive performance, while they often lack external validation, resulting in an overall high risk of bias. In the future, clinicians and nurses should learn from the evaluation criterions based on the PROBAST, then to test and verify them. Or machine learning algorithms usage is to construct models with operationalization and better predictive efficacy. [REGISTRATION: PROSPERO ID: CRD42024503844]