Predictive value of machine learning for the progression of gestational diabetes mellitus to type 2 diabetes: a systematic review and meta-analysis
Meng Zhao, Zhixin Yao, Yan Zhang, Lidan Ma, Wenquan Pang, Shuyin Ma, Yi-Jun Xu, Lili Wei
Abstract
BACKGROUND: This systematic review aims to explore the early predictive value of machine learning (ML) models for the progression of gestational diabetes mellitus (GDM) to type 2 diabetes mellitus (T2DM). METHODS: A comprehensive and systematic search was conducted in Pubmed, Cochrane, Embase, and Web of Science up to July 02, 2024. The quality of the studies included was assessed. The risk of bias was assessed through the prediction model risk of bias assessment tool and a graph was drawn accordingly. The meta-analysis was performed using Stata15.0. RESULTS: A total of 13 studies were included in the present review, involving 11,320 GDM patients and 22 ML models. The meta-analysis for ML models showed a pooled C-statistic of 0.82 (95% CI: 0.79 ~ 0.86), a pooled sensitivity of 0.76 (0.72 ~ 0.80), and a pooled specificity of 0.57 (0.50 ~ 0.65). CONCLUSION: ML has favorable diagnostic accuracy for the progression of GDM to T2DM. This provides evidence for the development of predictive tools with broader applicability.