Electrocardiographic recognition of left atrial enlargement

Abstract

The ECG is widely used as a screening test for left atrial enlargement (LAE). Surprisingly, the most widely used criterion of LAE, the P-terminal force in lead V1 (PTF-V1) has not been systematically evaluated to determine the optimal level of PTF-V1 for detection of LAE in clinical populations. Accordingly, we examined the relationship between PTF-V1 and left atrial size by echocardiogram in 361 patients and performed a Bayesian analysis of test performance in populations with a varying prevalence of LAE. As PTF-V1 increased from greater than or equal to 0.03 to greater than or equal to 0.08, sensitivity in the 82 patients with LAE (LA dimension greater than 40 mm) fell from 51% to 23%, and specificity rose from 70% to 93%. In our study population (LAE prevalence = 23%), diagnostic performance of criteria was: PTF-V1 greater than or equal to 0.03 greater than or equal to 0.04 greater than or equal to 0.05 greater than or equal to 0.06 greater than or equal to 0.08 Positive Predictive Accuracy 33 46 52 58 50 Negative Predictive Accuracy 83 83 84 83 80 Per Cent Correct Diagnosis 66 76 78 80 77 Positive predictive accuracy and per cent correct diagnosis improved progressively as PTF-V1 rose from greater than or equal to 0.03 to greater than or equal to 0.06, but fell at greater than or equal to 0.08. Applying our sensitivity and specificity data to Bayesian analysis, PTF-V1 greater than or equal to 0.06 performed best in all populations with prevalence of LAE less than or equal to 50%. We conclude that use of PTF-V1 greater than or equal to 0.06 is superior to the standard criterion of PTF-V1 greater than or equal to 0.04 for all purposes ranging from screening of a general population to evaluation of diseased individuals whose likelihood of LAE ranges up to 50%.