O40 Predictive models of individual risk of elective caesarean section complications: a systematic review

Abstract Introduction With increasing caesarean section (c-section) rates, personalised communication of risk has become paramount. A reliable tool to predict complications would support evidence-based discussion around planned mode birth. Method MEDLINE, Embase, Web Science, CINAHL, and the Cochrane Library were searched on 27th January with terms relating c-section, prognostic models, such as infection. Any study developing and/or validating a model for maternal c-section in English language after 1995 was included. Data extracted encompassed: source data, participant criteria, outcome be predicted, candidate predictors, actual sample size, development, performance. PROBAST (Prediction Risk Of Bias Assessment Tool) utilised bias analysis applicability concern studies. Result 7,752 studies identified, which 16 reviewed producing 3 where models identified predicted of: blood transfusion, spinal hypotension, postpartum haemorrhage. From studies, total 29 unique predictors 15 final model. Study authors deemed their exploratory, confirmatory respectively. None externally validated all had high due technique. Conclusion Few have been developed elective c-section. Existing predicting haemorrhage cannot recommended clinical practice. Future research should focus identifying known before surgery resulting models. Take-home Message Systematic review prediction C-section found none suitable