Prediction models for intrapartum fetal hypoxia: A systematic review

Background: Intrapartum fetal hypoxia (IFH) occurs infrequently and many cases occur unexpectedly, highlighting the importance of identifying at-risk fetuses to prevent hypoxic injury during birth. Objectives: This systematic review aims to evaluate existing prediction models for IFH and ascertain the predictor variables that were incorporated into most models. Search strategy: We searched MEDLINE and EMBASE databases from inception to December 2023 for synonyms of fetal distress, operative birth and neonatal asphyxia, with a filter for prediction models. Selection criteria: Prediction models for IFH were included. IFH was defined as evidence of hypoxia or operative intervention for suspected fetal distress. Data collection and analysis: Two authors independently extracted information on study design, outcome definition and validation. We evaluated study quality using the PROBAST tool. The number of models in which each predictor variable significantly contributed to the risk of IFH was quantified. Main results: From 874 articles, 23 publications reported on 41 eligible prediction models. 32 models reported on model development only, while seven models across three studies included both development and internal validation. Of these, two models were also externally validated and refit to the validation data, but investigated fetuses with antenatally-suspected growth restriction, limiting generalisability. Frequently reported predictor variables included parity (13 studies), cerebroplacental ratio (10 studies) and oxytocin use (6 studies). Conclusions: No prediction models for IFH applicable to the general population are currently ready for clinical application. External validation of existing models or development of new, reliable models applicable to all pregnant women is needed.