Discovering Five Types of Heart Failure with Artificial Intelligence Technology

Researchers at University College London (UCL) have identified five subtypes of heart failure that could predict a patient’s future risk, according to a study published in The Lancet Digital Health. The study analysed anonymised patient data from more than 300,000 patients aged over 30 diagnosed with heart failure in the UK over a 20-year period. Researchers used machine learning techniques to identify the five subtypes: early-onset, late-onset, metabolic, cardiometabolic and atrial fibrillation-related. Mortality risks varied widely by subtype, with corresponding 1-year mortality risks for all causes of 20% for early-onset, 46% for late-onset, 61% for atrial fibrillation-related, 11% for metabolic and 37% for cardiometabolic. The researchers have developed an app that clinicians can use to identify heart failure subtypes, potentially leading to more precise treatment strategies.