Machine learning to predict adverse perinatal outcomes: a systematic review and meta-analysis.
Machine learning doesn't outperform traditional statistics for predicting preterm birth and related complications, suggesting simpler approaches may suffice in obstetrics.
Meta-analysis of 90 studies finds ML prediction of preterm birth (AUC 0.73), SGA (AUC 0.68), and stillbirth (AUC 0.75) shows only moderate performance; ML is not significantly better than logistic regression for PTB/SGA (P=0.07/0.31); almost all studies had high risk of bias. This record was retained from the prior triage attempt for PubMed pipeline handoff.
What the study was
- Study design
- systematic_review_meta_analysis
- Category
- ai_ml_diagnostics
- Maturity
- Validated
- Journal
- EClinicalMedicine
Why it surfaced
EClinicalMedicine (Lancet group); PROSPERO-registered, high-quality methodology; important calibration signal for the AI/ML diagnostics field showing modest real-world performance; Oxford group with large dataset; T4 watchlist relevance.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.