Researchers analysed 81,362 Holter recordings from 54,395 individuals across 20 primary care centres in Israel, deriving atrial fibrillation burden, premature ventricular contraction burden and T-wave alternans and combining them with age in a random forest. Five-year discrimination was 0.75 for heart failure, 0.69 for ischaemic stroke and 0.79 for all-cause mortality, with the three markers adding up to 10% over age alone below age 75.
Why it is interesting: Tests whether routinely recorded single-lead ECG features add prognostic information beyond age in an unselected primary care population.