Is the diagnostic value of electrocardiography for the measurement of left ventricular mass index underestimated?

Dear editor,

I have read with great interest one of your recently published papers by Mohammad Noori etal. entitled: “Diagnostic Value of Electrocardiography Compared with Echocardiography in Measuring Left Ventricular Mass Index in Major Thalassemia Patients Over 10 Years of Age”.[1] In that article, having investigated the diagnostic value of electrocardiography (ECG) compared with echocardiography for the measurement of the left ventricular mass index (LVMI) in patients older than 10 years of age with thalassemia major, the authors concluded that ECG could not measure the LVMI accurately and, thus, should not be routinely used for this purpose.

By all accounts, the cornerstone of a diagnostic value study is the selection of a gold standard test, to which other procedures or tests of interest will be compared. In other words, the gold standard test will definitely confirm or rule out the outcome of interest. In the Noori et al. study, echocardiography was considered as the gold standard test. Although the high sensitivity and specificity of this procedure for the measurement of the LVMI are well-known, one must note that the gold standard for the measurement of the LVMI is surgery. In other words, echocardiography cannot definitely measure the LVMI inasmuch as it has some measurement errors compared to surgery. As a result, when we compare the diagnostic value of ECG with echocardiography for the measurement of the LVMI, the measurement errors of these two procedures will combine and compromise the diagnostic value of ECG.

Another feature of a diagnostic value study is its ability to determine a cut-off point in continuous scales for distinguishing the presence or absence of a specific outcome. This ability of diagnostic value studies is usually reported in articles as a receiver operating characteristic (ROC) curve. It seems that the authors should have reported the best cut-off point in the LVMI values (which was calculated with a formula mentioned in their study) for determining the presence of ventricular hypertrophy, which is a well-known risk factor for ventricular dysfunction.

All in all, this article is interesting and presents some valuable findings. The authors could increase the quality of their future studies by considering the points mentioned in this letter.