Researchprediction modelsfrailtydiabetesrisk stratification
Models Predict Frailty in Older Adults With Diabetes
7.1
A systematic review and meta-analysis published in J Med Internet Res (2026) evaluated models predicting physical and cognitive frailty in older adults with diabetes, searching six databases through December 2025 and including 24 studies (32 diagnostic models). Pooled discrimination was strong (AUC 0.851, 95% CI 0.820–0.882), logistic regression outperformed machine learning, but all studies had high risk of bias, limiting clinical utility.
Key Points
- 1Synthesizes 24 studies and 32 diagnostic models, reporting pooled AUC 0.851 and sensitivity 0.810.
- 2Demonstrates logistic regression outperforms machine learning (AUC 0.850 vs 0.785; P=.003), affecting model choice.
- 3Highlights pervasive high risk of bias and limited external validation, requiring prospective multicenter TRIPOD-adherent studies.
Scoring Rationale
Solid systematic synthesis with pooled metrics; constrained by high risk of bias and limited external validation.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
