Background: ECG criteria for detecting left ventricular hypertrophy (LVH) have been established and evaluated in populations with underrepresentation of Africans. We tested accuracy of LVH ECG criteria in a large population of black African adults with untreated, uncomplicated hypertension in comparison to LV mass determined with echocardiography. Design and method: In this subanalysis of the CoArtHA (Identifying most effective Treatment Strategies to Control Arterial Hypertension in sub-Saharan Africa) we analysed the value of different ECG criteria compared to cardiac geometry assessed by echocardiography at baseline. LVH was defined as left ventricular mass indexed to body surface area (LVMI) >95g/m 2 ; in women, >115g/m 2 ; in men. Results: In 1125 treatment naïve adults with uncomplicated hypertension (median age 53.4?years, 72?% women), echocardiographic LVH was present in 56 (5?%) participants. Strongest correlations with LVMI were observed for Cornell voltage product (rho?=?0.373) and Cornell voltage criteria (rho?=?0.327), all p?<?0.001. Standard cut offs yielded highest specificity for strain pattern (0.970) and Cornell product (0.923). Among continuous measures, the Cornell voltage product adjusted by +0.8?mV in women achieved the highest AUC (0.719), followed by its unadjusted form (AUC?0.651). Conclusions: The Cornell voltage product?particularly when adjusted by +0.8?mV in women?demonstrated the best overall diagnostic accuracy (AUC?0.719) for detecting LVH in this rural African hypertensive cohort, despite only moderate correlation with echocardiographic LVMI. Its implementation as a first line ECG criterion may help to stratify risk and prioritize patients for echocardiography in resource limited settings.
Nemtsova et al. (Fri,) studied this question.