Development and validation of nomograms for predicting adverse neonatal outcomes in women with polycystic ovary syndrome: a retrospective study.

In our study, we retrospectively enrolled 606 women with newly diagnosed polycystic ovary syndrome. Participants were divided into two cohorts: development cohort (n = 424) and validation cohort (n = 182). Multivariate logistic regression analyses were used to identify predictive indicators, and nomograms were developed and validated. We found that waist hip rate (WHR), testosterone levels, and fasting blood glucose (FBG) levels (WTF) could predict the small for gestational age; BMI, WHR and modified Ferriman-Gallwey Score (BWM) correlated with low Apgar scores; and BMI, WHR, modified Ferriman-Gallwey Score, testosterone levels, and FBG levels (BWMTF) correlated with adverse neonatal outcomes. The BWMTF nomogram was established, revealing perfect discrimination with the area under the receiver operating characteristic curve (AUC) and stratified five-fold cross-validation in development cohort (AUC = 0.75, Mean AUC = 0.75) and validation cohort (AUC = 0.68, Mean AUC = 0.75). Calibration plots showed good calibration. We established and validated three models for predicting adverse perinatal effects to guide preventive treatment protocols. Impact statementWhat is already known on this subject? Many studies have identified a large number of predictors, but also lack a comprehensively quantified tool to predict adverse neonatal outcomes in women with PCOS to guide the development of clinical treatment programs.What do the results of this study add? This article screened the high risks factors of adverse neonatal outcomes in women with PCOS, and three nomograms were established and validated. Also, the area under the receiver operating characteristic curve (AUC) and stratified five-fold cross-validation in development cohort and validation cohort showed good discrimination; Calibration plots showed good calibration.What are the implications of these findings for clinical practice and/or further research? Our scoring system could help clinicians evaluate these risks and conduct proper screening, prevention, and management to ameliorate the risk of neonatal disease in these patients.