Diabetes research and clinical practice · 2025
Development and validation of prediction models for gestational diabetes mellitus in first-trimester pregnant women.
СтатьяВалидационное исследованиеНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Liu Xiaohang, Hu Fang, Yan Ruohua, Liu Ruixia, Su Shaofei, Peng Yaguang, Peng Xiaoxia, Yin Chenghong
Аннотация
AIMS: Early intervention can reduce the risk of gestational diabetes mellitus (GDM) and related adverse pregnancy outcomes. Accordingly, this study aims to develop prediction models for GDM in early pregnant women.
METHODS: Pregnant women enrolled at Beijing Obstetrics and Gynecology Hospital were included in the derivation and temporal datasets. Women with electronic data capture records from other subcenters were included in the external validation dataset. Logistic regression and eXtreme Gradient Boosting (XGBoost) models were used to predict the risk of GDM.
RESULTS: A total of 20,435 pregnant women were included in the derivation dataset, with 1,997 pregnant women in the temporal validation dataset. Furthermore, 100 pregnant women were included in the external validation dataset. The logistic regression and XGBoost models demonstrated AUCs of 0.738 (95% CI: 0.707, 0.771) and 0.737 (95% CI: 0.706, 0.767) in temporal validation. For the external validation dataset, the AUCs were 0.674 (95% CI: 0.440, 0.879) and 0.737 (95% CI: 0.510, 0.929) for the logistic regression and XGBoost models, respectively.
CONCLUSION: Both the logistic regression and XGBoost models demonstrated satisfactory performance in the internal and temporal datasets. However, the XGBoost model showed more robust performance in the external validation dataset compared to the logistic regression model.
METHODS: Pregnant women enrolled at Beijing Obstetrics and Gynecology Hospital were included in the derivation and temporal datasets. Women with electronic data capture records from other subcenters were included in the external validation dataset. Logistic regression and eXtreme Gradient Boosting (XGBoost) models were used to predict the risk of GDM.
RESULTS: A total of 20,435 pregnant women were included in the derivation dataset, with 1,997 pregnant women in the temporal validation dataset. Furthermore, 100 pregnant women were included in the external validation dataset. The logistic regression and XGBoost models demonstrated AUCs of 0.738 (95% CI: 0.707, 0.771) and 0.737 (95% CI: 0.706, 0.767) in temporal validation. For the external validation dataset, the AUCs were 0.674 (95% CI: 0.440, 0.879) and 0.737 (95% CI: 0.510, 0.929) for the logistic regression and XGBoost models, respectively.
CONCLUSION: Both the logistic regression and XGBoost models demonstrated satisfactory performance in the internal and temporal datasets. However, the XGBoost model showed more robust performance in the external validation dataset compared to the logistic regression model.