A composite tool for predicting gestational diabetes or hypertensive disorders of pregnancy: development and temporal validation

AIMS: Cardiometabolic pregnancy complications, including gestational diabetes mellitus (GDM) and hypertensive disorders of pregnancy (HDP), share similar risk factors and increase cardiometabolic risk in women. Medical/lifestyle management reduces their risk but requires early identification of high risk individuals. This study aimed to develop a risk prediction model identifying women at high risk of developing GDM and/or HDP. METHODS: A composite prognostic risk prediction model for developing GDM and/or HDP comprising first trimester sociodemographic and clinical variables was developed using routinely collected health data from women delivering singleton pregnancies in Australia from 2016 to 2022. Variables comprised age; ethnicity; height; body mass index; parity; family history of diabetes; history of GDM, pre-eclampsia/eclampsia, chronic hypertension, polycystic ovary syndrome, chronic kidney disease and fetal macrosomia. Model development used logistic regression with performance evaluated using discrimination, calibration, bootstrapped internal and temporal validations, recalibration of intercept and slopes and decision curve analyses. RESULTS: GDM/HDP prevalence was 25.4% (n = 48,502). On temporal validation (n = 12,868), the model had a C-statistic of 0.739 (95% CI: 0.727, 0.738). Following recalibration, the calibration-in-the-large was -0.001 (95%CI: -0.04, 0.04) and the calibration slope was 1.00 (95% CI: 0.95, 1.05)]. On decision curve analysis, the model outperformed both treat-all and treat-none strategies across threshold probabilities from 8 to 72%. CONCLUSION: The model demonstrated fair-good discrimination and excellent calibration. Decision curve analysis indicated potential clinical utility across a range of threshold probabilities for identifying women at high risk. External validation is required.