Addressing systematic underdetection of left ventricular hypertrophy by guideline-recommended electrocardiographic criteria in women and individuals with overweight.
We aimed to probe the European Society of Cardiology (ESC) guideline recommended electrocardiographic (ECG) criteria for degree of under detection of left ventricular hypertrophy (LVH) across clinically relevant subgroups in a contemporary general population and if ECG-threshold recalibration could promote diagnostic equity.
We studied participants in the Copenhagen General Population Study. Cardiac computed tomography (CT) was used to define CT-LVH, and ECG-LVH was defined according to ESC guideline: Sokolow-Lyon, Cornell voltage, or Sokolow-aVL criteria. Diagnostic performance was assessed using Poisson regression across subgroups defined by age above 60 years, sex, overweight (BMI ≥ 25 kg/m2), hypertension, diabetes mellitus, obstructive pulmonary disease, and physical activity. ECG-thresholds were recalibrated to 95% specificity while maximizing sensitivity across all subgroups with disparities in both measures. Recalibration was performed in 50% of the population and validation in the remaining 50%.
We studied 9392 participants with median age of 60 years, of whom 58% were women. By CT 839 (9%) had CT-LVH. The guideline ECG-LVH definition had overall 20% sensitivity and 96% specificity. Sensitivity was lower in women than men (15% versus 26%; P < 0.001) and individuals with overweight compared to normal weight (16% versus 23%; P = 0.005). The recalibrated ECG-thresholds provided overall 36% sensitivity and 91% specificity in the validation cohort with consistent sensitivity and specificity across subgroups. The new recalibrated ECG criteria promoted diagnostic equity in LVH detection across sex and BMI subgroups, increasing overall sensitivity at a slight specificity cost.
We studied participants in the Copenhagen General Population Study. Cardiac computed tomography (CT) was used to define CT-LVH, and ECG-LVH was defined according to ESC guideline: Sokolow-Lyon, Cornell voltage, or Sokolow-aVL criteria. Diagnostic performance was assessed using Poisson regression across subgroups defined by age above 60 years, sex, overweight (BMI ≥ 25 kg/m2), hypertension, diabetes mellitus, obstructive pulmonary disease, and physical activity. ECG-thresholds were recalibrated to 95% specificity while maximizing sensitivity across all subgroups with disparities in both measures. Recalibration was performed in 50% of the population and validation in the remaining 50%.
We studied 9392 participants with median age of 60 years, of whom 58% were women. By CT 839 (9%) had CT-LVH. The guideline ECG-LVH definition had overall 20% sensitivity and 96% specificity. Sensitivity was lower in women than men (15% versus 26%; P < 0.001) and individuals with overweight compared to normal weight (16% versus 23%; P = 0.005). The recalibrated ECG-thresholds provided overall 36% sensitivity and 91% specificity in the validation cohort with consistent sensitivity and specificity across subgroups. The new recalibrated ECG criteria promoted diagnostic equity in LVH detection across sex and BMI subgroups, increasing overall sensitivity at a slight specificity cost.
Authors
Johansen Johansen, Kühl Kühl, Aspe Aspe, Lowes Lowes, Loft Loft, Pham Pham, Sigvardsen Sigvardsen, Nielsen Nielsen, Larsen Larsen, Graff Graff, Paulsen Paulsen, Kofoed Kofoed
View on Pubmed