Heart failure & cardiomyopathiesDiagnostic or prognostic study
Machine Learning Multicenter Risk Model to Predict Right Ventricular Failure After Mechanical Circulatory Support
- Prognostic model development and external validation study (hybrid prospective-retrospective multicentre cohort)
Authors' conclusion
Implementing routine clinical data, this multicenter cohort study derived and validated the STOP-RVF calculator as a personalized risk assessment tool for the prediction of RVF and RVF-associated all-cause mortality.
Study context
- Population
- Patients with advanced heart failure (HF) requiring continuous-flow left ventricular assist device (LVAD), April 2008 to July 2019
- Participants
- 798 vs 327
- Comparison
- STOP-RVF risk calculator vs published right ventricular failure (RVF) scores
- Derivationn = 798
- Validationn = 327
- Outcome
- RVF incidence (need for RV assist device or intravenous inotropes >14 days)
Performance
As reported in the abstract
RVF incidence and STOP-RVF discrimination
| Test or model | Value | STOP-RVF C statistic |
|---|---|---|
| Derivation | 24.2% | 0.75 (95% CI, 0.71-0.79) |
| Validation | 32.7% | 0.73 (95% CI, 0.67-0.80) |
Further findings
Comparison with published RVF scores
- STOP-RVF significantly better than Kormos et al score (C statistic)
- 0.58 (95% CI, 0.53-0.63)
- STOP-RVF significantly better than Drakos et al score (C statistic)
- 0.62 (95% CI, 0.57-0.67)
Intraoperative characteristics
- Inclusion of intraoperative characteristics did not improve model performance
Cumulative survival by risk group
- Cumulative survival higher in low-risk group (estimated <20% RVF risk) vs higher-risk groups
Original publication
Machine Learning Multicenter Risk Model to Predict Right Ventricular Failure After Mechanical Circulatory Support. JAMA Cardiol. Published 1 March 2024. doi:10.1001/jamacardio.2023.5372
Summarised from the published abstract; every number is checked against it