Performance of prediction models for incident heart failure and heart failure hospitalization in individuals with type 2 diabetes: a systematic review and meta-analysis
Authors' conclusion
While some HF and HFH models in diabetes show acceptable/good long-term discriminative performance, the scarcity of utility analyses and prospective testing limits translation to clinical settings.
Study context
- Population
- Individuals with T2DM65 studies
- Comparison
- RECODe vs DMRS vs WATCH-DM 2022 (r) vs WATCH-DM 2022(i) vs DM-CURE
- Outcome
- Discrimination of multivariable models predicting incident HF and HF hospitalization (HFH)
Methodological detailsBayesian meta-analysis, PROBAST risk of bias
- Bayesian meta-analysis
- PROBAST risk of bias
Pooled result
As reported in the abstract
Pooled discrimination (models validated in ≥3 cohorts, excluding high risk of bias)
RECODe
0.711
95% CI 0.651–0.767
DMRS
0.758
95% CI 0.684–0.827
WATCH-DM 2022 (r)
0.705
95% CI 0.624–0.795
WATCH-DM 2022(i)
0.718
95% CI 0.587–0.850
DM-CURE
0.837
95% CI 0.757–0.913
Further pooled results
Models included
- Heart failure (HF) prediction models
- 56
- HFH prediction models
- 44
Prediction horizon, prospective testing and clinical utility
- Well-performing models reporting a prediction horizon <10 years
- 0
- Models prospectively tested
- 0
- Models with clinical utility analysis
- 3
Original publication
Performance of prediction models for incident heart failure and heart failure hospitalization in individuals with type 2 diabetes: a systematic review and meta-analysis. Eur J Heart Fail. Published 10 May 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →