Widely accessible prognostication using medical history for fetal growth restriction and small for gestational age in nationwide insured women

Abstract

Objectives: Prevention of fetal growth restriction/small for gestational age is adequate if screening is accurate. Ultrasound and biomarkers can achieve this goal; however, both are often inaccessible. This study aimed to develop, validate, and deploy a prognostic prediction model for screening fetal growth restriction/small for gestational age using only medical history. Methods: From a nationwide health insurance database (n=1,697,452), we retrospectively selected visits of 12-to-55-year-old females to 22,024 healthcare providers of primary, secondary, and tertiary care. This study used machine learning (including deep learning) to develop prediction models using 54 medical-history predictors. After evaluating model calibration, clinical utility, and explainability, we selected the best by discrimination ability. We also externally validated and compared the models with those from previous studies, which were rigorously selected by a systematic review of Pubmed, Scopus, and Web of Science. Results: We selected 169,746 subjects with 507,319 visits for predictive modeling. The best prediction model was a deep-insight visible neural network. It had an area under the receiver operating characteristics curve of 0.742 (95% confidence interval 0.734 to 0.750) and a sensitivity of 49.09% (95% confidence interval 47.60% to 50.58% using a threshold with 95% specificity). The model was competitive against the previous models in a systematic review of 30 eligible studies of 381 records, including those using either ultrasound or biomarker measurements. We deployed a web application to apply the model. Conclusions: Our model used only medical history to improve accessibility for fetal growth restriction/small for gestational age screening. However, future studies are warranted to evaluate if this model's usage impacts patient outcomes.