Heart failure & cardiomyopathies
CONFIDENT-HFpEF: a machine learning-based risk stratification for mortality and hospitalization using multimodal real-world data
Authors' conclusion
CONFIDENT prognostic models for all-cause mortality and HF hospitalization using routinely collected variables can reliably predict outcomes and potentially facilitate personalized care and trial recruitment strategies in HFpEF.
Study context
- Population
- Patients with heart failure with preserved ejection fraction (HFpEF) according to the HFA-PEFF criteria, with ≥ 2 years of follow-up, included from 2013 to 2022patients included in the study
- Participants
- 1,208 patients
- Predictors
- Machine learning-based prognostic models using electronic health records, lab tests, echocardiography, and electrocardiography (routinely collected variables)
- Predicted outcome
- All-cause mortality and heart failure (HF) hospitalization
Methodological detailsthree centres in Europe and the USA, machine learning-based prognostic models, external validation cohort
- Three centres in Europe and the USA
- Machine learning-based prognostic models
- External validation cohort
Model performance
As reported in the abstract
C-index
CONFIDENT model, first training cohort
0.68(95% CI 0.62-0.74)
CONFIDENT model, second training cohort
0.71(95% CI 0.64-0.78)
CONFIDENT model, validation cohort
0.72(95% CI 0.65-0.78)
PREDICT-HFpEF score, first training cohort
0.66(95% CI 0.54-0.72)
PREDICT-HFpEF score, second training cohort
0.65,(95% CI 0.55-0.72)
PREDICT-HFpEF score, validation cohort
0.67(95% CI 0.59-0.73)
| Test or model | P-value vs CONFIDENT model |
|---|---|
| PREDICT-HFpEF score, first training cohort | P-value = .006 |
| PREDICT-HFpEF score, second training cohort | P-value < .001 |
| PREDICT-HFpEF score, validation cohort | P-value = .036 |
P-value vs CONFIDENT model
PREDICT-HFpEF score, first training cohortP-value = .006
PREDICT-HFpEF score, second training cohortP-value < .001
PREDICT-HFpEF score, validation cohortP-value = .036
Further findings
All-cause mortality, compared with MAGGIC
- Similar results observed when compared to the Meta-Analysis Global Group In Chronic Heart Failure Risk Score (MAGGIC)
Heart failure (HF) hospitalization
- HF hospitalization model outperformed both comparators, including MAGGIC+ natriuretic peptide
Baseline characteristics
- Mean age
- 72 ± 12
- Mean body mass index
- 32 ± 9 kg/m²
Risk of HF hospitalization and all-cause mortality · 2 years
- HF hospitalization, range
- 13 to 44%
- All-cause mortality, range
- 9 to 19%
Original publication
CONFIDENT-HFpEF: a machine learning-based risk stratification for mortality and hospitalization using multimodal real-world data. ESC Heart Fail. Published 7 April 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →