Objective: To assess racial differences in the accuracy of standard electrocardiographic (ECG) criteria in the diagnosis of left ventricular hypertrophy (LVH).
Design: The sensitivity and specificity of standard ECG criteria were compared in blacks and whites using echocardiographic LVH as the reference standard.
Setting: Eight worksite-based hypertension clinics in New York, NY.
Patients: A sample of 122 black and 148 white hypertensive patients.
Results: The prevalence of ECG-LVH was two to six times higher in blacks than in whites, depending on the criteria used (range, 6% to 24% in blacks vs 1% to 7% in whites; P = .0005 to .19 for black-white comparisons). The difference in prevalence of echocardiographic LVH [corrected], however, was less striking and did not attain statistical significance (26% in blacks and 20% in whites; P greater than .2). The sensitivity of the ECG was low (range, 3% to 17%) and did not differ significantly between the two races for any of the conventional criteria; specificity, however, was lower in blacks for all criteria (range, 73% to 94% vs 95% to 100% for whites; P = .0001 to .09). The predictive value of a positive ECG was consistently, although not significantly, lower in the black subjects. Black race was the strongest independent predictor of decreased ECG specificity in multiple logistic regression analysis that also considered age, gender, body mass index, left ventricular mass index, and smoking.
Conclusions: Commonly used ECG criteria for the detection of LVH have a poor sensitivity in both black and white hypertensives and a lower specificity in blacks than in whites; this may lead to a greater number of false-positive diagnoses in black patients, as well as to an overestimation of black-white difference in LVH prevalence.