Objective:

To evaluate the effectiveness of artificial intelligence in predicting the transition from gestational diabetes mellitus (GDM) to type 2 diabetes mellitus (T2DM).

Approach:

• AI Techniques: The study reviews various AI methods, including random forest, decision tree, logistic regression, multilayer perceptron, naïve Bayes, and extreme gradient boosting, for predicting patient outcomes.
• Data Analysis: AI algorithms analyze comprehensive patient data to identify patterns and correlations that improve prediction accuracy compared to traditional statistical methods.

Key Findings:

• Women with a history of GDM have a tenfold higher risk of developing T2DM compared to those with normoglycemic pregnancies [Vounzoulaki E, Khunti K, Abner SC, Tan BK, Davies MJ, Gillies CL. Progression to type 2 diabetes in women with a known history of gestational diabetes: systematic review and meta-analysis. BMJ. May 13, 2020;369:m1361.].
• 30%-50% of women with previous GDM develop T2DM within 5 to 10 years postpartum [Ferrara A. Increasing prevalence of gestational diabetes mellitus: a public health perspective. Diabetes Care. Jul 2007;30 Suppl 2:S141-S146.].
• AI models demonstrate superior predictive performance for risk stratification compared to traditional regression models.

Interpretation:

The increasing prevalence of GDM necessitates effective tools for risk stratification to improve screening and intervention strategies.

Limitations:

• Postpartum screening for T2DM remains suboptimal, with compliance rates as low as 16%-19% [Shah BR, Lipscombe LL, Feig DS, Lowe JM. Missed opportunities for type 2 diabetes testing following gestational diabetes: a population-based cohort study. BJOG. Nov 2011;118(12):1484-1490.].
• Logistical difficulties and patient perceptions contribute to poor adherence to screening guidelines [Shah BR, Lipscombe LL, Feig DS, Lowe JM. Missed opportunities for type 2 diabetes testing following gestational diabetes: a population-based cohort study. BJOG. Nov 2011;118(12):1484-1490.].

Conclusion:

AI has the potential to enhance predictive accuracy for the transition from GDM to T2DM.