ECG does not rule out LV hypertrophy

ECG does not rule out LV hypertrophy


By Caroline Price



07 September 2007



Br Med J 2007; Advance online publication


MedWire News: Electrocardiographic criteria should not be used to rule out left ventricular (LV) hypertrophy in patients with hypertension, say clinicians in an advance online publication by the British Medical Journal.


Matthias Egger (Universities of Bern, Switzerland, and Bristol, UK) and colleagues conclude this after conducting a systematic review of studies testing the accuracy of six different electrocardiographic indexes.


Accurate and early diagnosis of LV hypertrophy is an important component of the care of hypertension patients, in whom it leads to a five- to 10-fold increase in cardiovascular risk, explain the researchers, but the appropriate diagnostic work-up of suspected LV hypertrophy remains unclear.


Egger and co-workers set out to clarify the accuracy of commonly used electrocardiographic indexes, focusing on their ability to rule out LV hypertrophy in patients with arterial hypertension.


"As the electrocardiogram (ECG) will mainly be used to rule out the diagnosis of LV hypertrophy, we were particularly interested in the sensitivity and the likelihood ratio of a negative ECG result," they note.


The team searched electronic databases, reference lists of relevant studies and reviews, and discussions with experts for observational studies that evaluated the accuracy of electrocardiographic indexes for LV hypertrophy diagnosis, and established the presence of LV hypertrophy using echocardiography.


This yielded 21 observational studies, involving a total of 5608 patients, 10 of which were conducted in primary care and 11 in secondary care.


The median prevalence of LV hypertrophy was 33% in primary care settings, and 65% in secondary care.


The researchers focused their analysis on data for the six most commonly used indexes: the Sokolow-Lyon voltage index; the Cornell voltage and Cornell product indexes; the Gubner index; and the Romhilt-Estes score with two different thresholds.


The results showed that the median sensitivity ranged from 10.5% for the Gubner index to 21% for the Sokolow-Lyon index.


The median negative likelihood ratio was similar across these different indexes, ranging from 0.85 for the Romhilt-Estes score (with a threshold for positive test of ≥4 points) to 0.91 for the Gubner index.


Using the median likelihood ratio from the Romhilt-Estes score (four points) in primary care, a negative ECG result would reduce the typical pre-test probability of 33% to 31%, Egger et al report.


In secondary care, the typical pre-test probability of 65% would be reduced to 63%.


"Irrespective of the index used, the ECG is a poor screening tool to exclude LV hypertrophy in hypertensive patients in primary and secondary care settings," the authors write.


Journal