Heart failure & cardiomyopathies
Evaluating the Performance and Potential Bias of Predictive Models for Detection of Transthyretin Cardiac Amyloidosis
Authors' conclusion
Deep learning, echo-based models to detect ATTR-CM demonstrated best overall discrimination when compared to 2 other models in external validation with low risk of harms due to racial bias.
Study context
- Population
- Patients with heart failure in an integrated health system, with ATTR-CM or matched heart failure controls176 confirmed cases of ATTR-CM and 3,192 control patients · confirmed cases of ATTR-CM 176 · control patients 3,192
- Index test or model
- 4 algorithms for ATTR-CM detection compared against confirmed ATTR-CM
- Target condition
- transthyretin amyloid cardiomyopathy (ATTR-CM)
Methodological detailsage- and sex-matched controls, integrated health system, 2010 to 2022, target 5% prevalence
- Age- and sex-matched controls
- Integrated health system
- 2010 to 2022
- Target 5% prevalence
Performance
As reported in the abstract
| Test or model | AUC |
|---|---|
| Huda et al model | 0.49 |
| EchoNet-LVH | 0.88 |
| EchoGo Amyloidosis | 0.92 |
| Mayo ATTR-CM Score | 0.79 |
AUC
Huda et al model0.49
EchoNet-LVH0.88
EchoGo Amyloidosis0.92
Mayo ATTR-CM Score0.79
Further findings
Self-identified race of analytical cohort
- White
- 79.2%
- Black
- 9.0%
Bias auditing (fairness metrics)
- Met fairness criteria for equal opportunity among patients who identified as Black
Original publication
Evaluating the Performance and Potential Bias of Predictive Models for Detection of Transthyretin Cardiac Amyloidosis. JACC Adv. Published 5 July 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →