Left ventricular hypertrophy in the automatic electrocardiogram report and mortality risk

Area of Science:

• Cardiology
• Digital Health
• Public Health

Background:

• Left Ventricular Hypertrophy (LVH) causes myocardial changes detectable by ECG, but current criteria have low sensitivity compared to echocardiography.
• Automated ECG analysis offers large-scale potential, yet studies on the prognostic value of automatically detected LVH are limited.

Purpose of the Study:

• To assess the association between the Glasgow program's LVH score from ECGs and overall mortality in a Brazilian primary care population.
• To determine if automated ECG-derived LVH is a predictor of mortality.

Main Methods:

• Analysis of digital ECGs from the CODE cohort (2010-2017) in patients over 18 years old.
• Utilized the University of Glasgow Automated ECG Interpretation program for LVH scoring (definite ≥6.3, probable 5.0-6.3, possible 4.0-5.0).
• Cox regression analysis, adjusted for age, sex, and comorbidities, evaluated the relationship between LVH score and mortality.

Main Results:

• The study included over 1.3 million patients, with an 18.5% prevalence of abnormal LVH-Glasgow scores (≥4.0).
• During a median follow-up of 3.47 years, the all-cause mortality rate was 2.68%.
• Definite LVH (HR 1.64) and probable LVH (HR 1.18) showed significantly higher risks of overall mortality, with possible LVH also associated with increased risk (HR 1.09).

Conclusions:

• The LVH-Glasgow score derived from automated ECG analysis serves as a valuable prognostic tool.
• Elevated LVH-Glasgow scores are independently associated with increased all-cause mortality in this primary care cohort.