On January 2, 2024, JAMA Cardiology published a groundbreaking study conducted by researchers at the National Heart and Lung Institute, Imperial College London. Led by Arunashis Sau, PhD, and Joseph Barker, MRes, the team developed an innovative artificial intelligence model named AIRE-HTN. This model enhances the prediction of hypertension and related cardiovascular risks. Supported by Imperial's British Heart Foundation Centre for Excellence Award and other organizations, the study utilized 65,610 ECGs from a UK-based volunteer cohort to validate AIRE-HTN's efficacy.
The AIRE-HTN model was initially trained on a vast derivation cohort involving 1,163,401 ECGs from 189,539 patients at the Beth Israel Deaconess Medical Center in Boston. The derivation cohort, with a mean age of 57.7 years and comprising 52.1% women and 64.5% White individuals, provided a robust foundation for the model. In contrast, the UK-based cohort had a mean age of 65.4 years, with 51.5% women and 96.3% White individuals.
AIRE-HTN has proven to be an independent predictor of several severe health conditions, including cardiovascular death, heart failure, myocardial infarction, ischemic stroke, and chronic kidney disease. The model demonstrated a notable ability to predict incident hypertension with a C-index of 0.70 in both the derivation and UK cohorts. Notably, individuals in the highest quartile of AIRE-HTN scores exhibited a fourfold increased risk for developing hypertension.
The model's predictive power was significantly additive to traditional clinical markers. It achieved a continuous net reclassification index of 0.44 for the medical center cohort and 0.32 for the UK cohort. The study meticulously evaluated incident hypertension in 19,423 individuals from the medical center cohort and 35,806 individuals from the UK cohort who did not have hypertension at baseline.
While the study defined hypertension using International Classification of Diseases codes, which may lack granularity compared to contemporary guidelines, it still highlighted AIRE-HTN's potential to enhance surveillance programs and primordial prevention strategies.
“Results of exploratory and phenotypic analyses suggest the biological plausibility of these findings. Enhanced predictability could influence surveillance programs and primordial prevention,”
- Arunashis Sau, PhD, and Joseph Barker, MRes, National Heart and Lung Institute, Imperial College London, England.
Leave a Reply