bmj.com Rapid Responses for Pewsner et al., 335 (7622) 711

Register for free services | Subscribe | Sign In

Printer-friendly Page
RSS Feeds

Rapid Responses to:

RESEARCH:
Daniel Pewsner, Peter Jüni, Matthias Egger, Markus Battaglia, Johan Sundström, and Lucas M Bachmann
Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review
BMJ 2007; 335: 711 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

EKG in Left Ventricular Hypertrophy: Sameer Chadha, Shikha Mehta, Sumeet Chadha Medical Students (4 September 2007)

QRS voltage criteria for LVH can be useful: Peter J Bourdillon (10 September 2007)

Re: QRS voltage criteria for LVH can be useful: Khaled Alfakih (16 September 2007)

Ethnic variations in the ECG: Andrew P Vanezis, Raj Bhopal (7 October 2007)

Correction to response posted on October 07 by Vanezis and Bhopal: Andrew Peter Vanezis, Raj Bhopal (10 October 2007)

Left ventricular hypertrophy and QT dispersion in hypertensive patients: Antoni Sis� Almirall, Antoni Dalf� Baqu�. (11 October 2007)

EKG in Left Ventricular Hypertrophy 4 September 2007

Sameer Chadha,
Medical Student
Maulana Azad Medical College, New Delhi, India,
Shikha Mehta, Sumeet Chadha Medical Students
Send response to journal:
Re: EKG in Left Ventricular Hypertrophy

Simplified Criteria for Diagnosing LVH
1)Deepest S wave in lead V1 or V2, plus tallest R wave in lead V5 or V6 > 35 and/or R wave in lead aVL > 12.
2)Patient > 35 years old.
3)Left ventricular (LV) "Strain".
Additional Voltage Criteria may occasionally be needed to diagnose LVH.
1)An R wave > 20 in any inferior lead (II, III, or aVF).
2)A deep S wave ( > 20-25) in lead V1 or lead V2.
3)A tall R wave ( > 25) in lead V5.
4)A tall R wave ( > 20) in lead V6.
Strain is a pattern of asymmetric ST segment depression and T wave inversion. LV strain is most commonly seen in one or more leads that look at the left ventricle (leads I, aVL, V4, V5, V6); less commonly it can be seen in inferior leads.
If a strain equivalent pattern occurs in association with voltage for LVH, specificity for true LVH is greatly enhanced compared to the voltage criteria alone.
Competing interests: None declared

QRS voltage criteria for LVH can be useful 10 September 2007

Peter J Bourdillon,
Honorary Senior Lecturer
ECG Department, Hammersmith Hospital, London W12 0HS
Send response to journal:
Re: QRS voltage criteria for LVH can be useful

While I do not want to pretend that the ECG is a good screening test for left ventricular hypertrophy (LVH), Pewsner et al ¹ do not portray the investigation in its best light. Their analysis takes no account of age or race, only partly takes account of gender, and does not combine criteria for the diagnosis of LVH.

It has long been known that age, gender and race are important determinants of the ECG ². Ignoring them leads to reduced specificity and results in �disease of electrocardiographic origin� in those having the test for screening purposes. In particular, the upper normal limits of QRS voltages in black males are greater than in white males, while the difference between black females and white females increases as the subjects become older ³.

Since 2002 at the Hammersmith Hospital we have been routinely reporting the upper limits of normal QRS voltage for R_aVL, R_aVL + S_V3 and R_V5 + S_V1 on every General Practitioner referral for an ECG, the request form of which states hypertension. These upper normal limits are calculated as the mean plus 2 standard deviations from the tables in Rautaharju et al ³. If any one of the 3 measurements is greater than the age-, gender- and race- adjusted upper limit of normal a diagnosis of LVH is suggested. The table shows the number of LVH diagnoses made in 1638 consecutive referrals over a 45 month period (850 female; white 921, black 418, from the Indian- subcontinent 171, other 128; age range 25 � 74 years, median 56) using both the Rautaharju-derived criteria and the criteria in Pewsner�s paper. The amplitude measurements were obtained from the GE Marquette 12SL program.

ECG measurement

Hammersmith criteria

Pewsner criteria

Sokolow-Lyon *

72

173

R_aVL

200

�

Cornell

96

188

Cornell product

�

272

Gubner

�

114

Any criterion positive

302 (18.4%)

440 (26.9%)

* the Hammersmith criteria only use R_V5 + S_V1; of the 173 that were positive with the Pewsner criteria, R_V6 was taller than R_V5 in 11

The table shows firstly that the Pewsner criteria result in roughly twice as many diagnoses of LVH using the Sokolow-Lyon and Cornell measurements than the criteria derived from Rautaharju�s study, and secondly that if the diagnosis of LVH only requires that any one of the criteria to be positive, there is a substantial increase in the number of diagnoses of LVH. The table also shows the value of using R_aVL alone.

The result of requiring only one of three criteria to be positive to make the diagnosis of LVH, as in the case of the Hammersmith criteria, is an increase in the false positive rate. The author knows of no data in the literature of the effect of combining age-, gender- and race-adjusted ECG measurements. Theoretically three independent tests that each are normally distributed, i.e. having 2.5% of measurements above the upper limit of normal, will result in a specificity of 92.7% (0.975³*100). However, ECG measurements are not independent, so that the actual specificity will be somewhat higher. Table 2 of the Pewsner paper reports the median specificity for Sokolow- Lyon of 89%, Cornell 96%, Cornell product between 85 and 97% and Gubner 96%. If these 4 measurements were to be combined, the theoretical specificity would be about 75%; or 82% (0.89*0.96*0.96*100) if the Cornell product is omitted. Even allowing for the correlation between the ECG measurements, the resulting specificity would be too low for a screening test.

Another concern I have is the (conventional) use of the echocardiogram as the determinant of LVH. Sundstr�m et al ⁴ found in 22 males who had 2 echocardiograms in a month that the coefficient of variation of left ventricular mass index was 12.5%. This is comparable to the coefficient of variation of Sokolow-Lyon 10.0%, Gubner 13.7% and Cornell 10.3% found in 77 subjects with hypertension recorded again after a week by van den Hoogen et al ⁵. Further, using magnetic resonance imaging (MRI) as the gold standard in 32 subjects with hypertension, Alfakih et al ⁶ showed that the intra-observer variability in measurements of 2D echocardiographic left ventricular mass, together with the wide limits of agreement with MRI, were sufficiently large to make serial estimates unreliable. The implication is that if 1638 people with hypertension were to have ECGs and echocardiograms on two occasions a month apart there would be significant differences in LVH classification; and that a doctor requesting the tests would not know whether the first, the second or neither classification were the correct one.

The reasons the Hammersmith Hospital ECG department decided to use Rautaharju�s data are firstly, that no previous study was found that provided age- and gender- adjusted normal limits for black males and females living in a westernised society, and secondly, it was the first study to report normal limits on subjects between the ages of 65 and 74. To use MRI as a screening test for LVH is clearly impracticable. The alternative of combining 3 age-, gender- and race-adjusted ECG measurements to generate a test with a relatively low sensitivity but a high specificity is a pragmatic one.

1. Pewsner D, J�ni P, Egger M, Battaglia M, Sundstr�m J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. doi:10.1136/bmj.39276.636354.AE

2. Simonson E. Differentiation between Normal and Abnormal in Electrocardiography. CV Mosby Company, St Louis, 1961

3. Rautaharju PM, Zhou SH, Calhoun HP. Ethnic differences in ECG amplitudes in North American white, black, and Hispanic men and women. Effect of obesity and age. J Electrocardiol 1994; 27 (suppl):20-31

4. Sundstr�m J, Lind L, �rnl�v J, Zethelius B, Andr�n B, Lithell HO. Echocardiographic and Electrocardiographic Diagnoses of Left Ventricular Hypertrophy Predict Mortality Independently of Each Other in a Population of Elderly Men. Circulation 2001;103;2346- 2351

5. van den Hoogen JPH, Mol WH, Kowsoleea A, van Ree JW, Thien T, van Weel C. Reproducibility of electrocardiographic criteria for left ventricular hypertrophy in hypertensive patients in general practice. Eur Ht J 1992;13:1606-1610

6. Alfakih K, Bloomer T, Bainbridge S, Bainbridge G, Ridgway J, Williams G, Sivananthan M. A comparison of left ventricular mass between two-dimensional echocardiography using fundamental and tissue harmonic imaging, and cardiac MRI in patients with hypertension. Eur J Radiol. 2004;52:103-9

Competing interests: None declared

Re: QRS voltage criteria for LVH can be useful 16 September 2007

Khaled Alfakih,
SpR in Cardiology
City Hospital, Nottingham, NG5 1PB
Send response to journal:
Re: Re: QRS voltage criteria for LVH can be useful

Assessment of LVH in Hypertension
The ECG is an economical initial test for the assessment of LVH. The systematic review by Pewsner et al. 1 confirms what is known about the various ECG criteria for the detection of LVH, which is that while they are highly specific, they are not very sensitive. ECG criteria identify a group of patients at the severe end of the LVH spectrum when compared to the left ventricle mass (LVM) as measured by cardiac MRI. 2 This is important because of the incremental relationship between LVM and cardiovascular risk. 3 Hence patients who are found to have LVH on ECG criteria are in a high risk group and should have their BP treated to the lower targets recommended by the hypertension societies, and their cardiovascular risk factors addressed. The European Society of Hypertension recommends the use of the Sokolow�Lyon voltage and the Cornell product criteria for the assessment of LVH, as they were used in the LIFE study, which confirmed that antihypertensive agents, that reduce LVH, beyond lowering of BP, confer further reduction in morbidity and mortality. 4
In view of the low sensitivity of the ECG criteria of LVH, hypertensive patients who do not have ECG evidence of LVH should be screened further. The ideal tool for this in terms of accuracy, availability and cost is echocardiography. However, M-mode LVM calculation should not be performed as it is not reproducible. Expert centres calculate that the LVM would need to change by more than 18% of its initial value for it to be detected by the same observer. 5 This is mainly because the cubing of linear measurements in the LVM formula exaggerates any errors. Furthermore, when establishing normal ranges, the M-mode LVM cube function formula results in high standard deviations, and hence underestimation of the prevalence of LVH when upper limits of normal are calculated. 6 Instead simple measurements of wall thickness and LV radius using M-mode, or 2-D measurements, gives a more accurate and reproducible assessment of the degree and the geometric type of LVH and provides more prognostic information than the calculated LVM. 7
1. Pewsner D, J�ni P, Egger M, Battaglia M, Sundstr�m J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. doi:10.1136/bmj.39276.636354.AE
2. Alfakih K, Walters K, Jones T, Ridgway J, Hall AS, Sivananthan M. New gender-specific partition values for ECG criteria of left ventricular hypertrophy: recalibration against cardiac MRI. Hypertension 2004; 44:175�179.
3. Schillaci G, Verdecchia P, Porcellati C, Cuccurullo O, Cosco C, Perticone F. Continuous relation between left ventricular mass and cardiovascular risk in essential hypertension. Hypertension 2000; 35:580�586.
4. Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, Faire U, et al. Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 2002; 359:995�1003.
5. de Simone G, Muiesan ML, Ganau A, Longhini C, Verdecchia P, Palmieri V, et al. Reliability and limitations of echocardiographic measurement of left ventricular mass for risk stratification and follow-up in single patients: the RES trial. Working Group on Heart and Hypertension of the Italian Society of Hypertension. Reliability of M-mode Echocardiographic Studies. J Hypertens 1999; 17:1955�1963
6. Korner PI, Jennings GL. Assessment of prevalence of left ventricular hypertrophy in hypertension. J Hypertens 1998; 16:715�723.
7. Muiesan ML, Salvetti M, Monteduro C, Bonzi B, Paini A, Viola S, et al. Left ventricular concentric geometry during treatment adversely affects cardiovascular prognosis in hypertensive patients. Hypertension 2004; 43:731�738.
Competing interests: None declared

Ethnic variations in the ECG 7 October 2007

Andrew P Vanezis,
Foundation Doctor
Colchester General Hospital, CO4 5JL,
Raj Bhopal
Send response to journal:
Re: Ethnic variations in the ECG

Rapid response article to Pewsner et al.(1)
This article does a good job at highlighting the danger of using the ECG as a sole diagnostic tool for detecting left ventricular hypertrophy (LVH) and subsequent cardiovascular risk status. Its main conclusions are that the current commonly used ECG criteria should not be used to rule out LVH i.e. the sensitivity of the tests are poor. Furthermore the paper shows that none of the more modern criteria are superior to the Sokolow- Lyon criteria which was devised almost 60 years ago and defines LVH as a total voltage in SV1 + RV5/6 of greater than 3.5 mV.(2)
However we agree with Bourdillon et al assessments that this article has not stratified the data to take into account age, gender and ethnicity.(3) Referring specifically to ethnicity, the literature shows that ethnic variation influences the accuracy of ECG criteria used to detect LVH is significant.
A recent review undertaken by the authors identified five suitable studies in the literature which looked at ethnicity (specifically African origin and White population groups) and how this factor affects the sensitivity and specificity of the ECG (using the Sokolow-Lyon and Cornell criteria) for detecting LVH.(4-8)
The main finding of our review was that the Sokolow-Lyon criteria are less specific for LVH in African origin people than in White people. When the results are taken as a whole they mirror the results obtained by Pewsner et al i.e. the specificity for these tests are good but the sensitivity is poor. However there were slightly higher sensitivity values for the African origin group (Cornell: 31.2%; Sokolow-Lyon: 32.9%) compared to the White group (Cornell: 26.5%; Sokolow-Lyon: 18.2%).
Specificity was high using both criteria in the White groups (Cornell: 87.4%; Sokolow-Lyon: 88.9%) but was much lower in the African origin group using the Sokolow-Lyon criteria (72.1%). Specificity was high however for the African origin group using the Cornell criteria (86.2%).
Pewsner et al shows that the Sokolow-Lyon index remain the best ECG criteria for the detection of LVH in hypertensive patients when taken as a whole. When contrasting the diagnostic accuracy of the Cornell and Sokolow -Lyon criteria across different ethnic groups, our evidence favours the Cornell criteria over the Sokolow-Lyon criteria in research and service contexts. However overall we agree with Pewsner et al that the ECG is not sufficient and would like to add to that the fact that it does not have equivalent validity across ethnic groups, which is important in the modern world where the mixing of different ethnic population groups calls for clarity in this issue.
1. Pewsner D, J�ni P, Egger M, Battaglia M, Sundstr�m J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. doi:10.1136/bmj.39276.636354.AE2.
2. Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limb leads. Am Heart J 1949;37:161-863.
3. Peter J Bourdillon. QRS voltage criteria for LVH can be useful. Response to Pewsner D, J�ni P, Egger M, Battaglia M, Sundstr�m J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. doi:10.1136/bmj.39276.636354.AE
4. Okin PM, Wright JT, Nieminen MS, Jern S, Taylor AL, Phillips R, et al. Ethnic differences in electrocardiographic criteria for left ventricular hypertrophy: the LIFE study. American Journal of Hypertension 2002;15(8):663-671.
5. Chapman JN, Mayet J, Chang CL, Foale RA, Thom SAM, Poulter NR. Ethnic differences in the identification of left ventricular hypertrophy in the hypertensive patient. American Journal of Hypertension 1999;12(5):437.
6. Rautaharju PM MT, Siscovick D, Zhou SH, Gardin JM, Kronmal R, Furberg CD, Borhani NO, Newman AB. The CHS Collaborative Research Group. Classification accuracy of electrocardiographic criteria for left ventricular hypertrophy in normal weight and overweight older adults. Annals of Noninvasive Electrocardiology 1996;1:121-132.
7. Crow RS, Prineas RJ, Rautaharju P, Hannan P, Liebson PR. Relation Between Electrocardiography and Echocardiography for Left Ventricular Mass in Mild Systemic Hypertension (Results from Treatment of Mild Hypertension study). The American Journal of Cardiology 1995;75(17):1233.
8. Lee DK, Marantz PR, Devereux RB, Kligfield P, Alderman MH. Left ventricular hypertrophy in black and white hypertensives. Standard electrocardiographic criteria overestimate racial differences in prevalence. JAMA 1992;267(24):3294-3299.
Competing interests: None declared

Correction to response posted on October 07 by Vanezis and Bhopal 10 October 2007

Andrew Peter Vanezis,
Foundation Doctor
Colchester General Hospital, CO4 5JL,
Raj Bhopal
Send response to journal:
Re: Correction to response posted on October 07 by Vanezis and Bhopal

Rapid response article to Pewsner et al.1
Ethnicity is relevant to judging ECG criteria for left ventricular hypertrophy
Pewsner et al highlight the danger of using the ECG for detecting left ventricular hypertrophy (LVH), particularly as the sensitivity of the tests is poor. They conclude that no criteria are superior to the Sokolow -Lyon criteria.2 Our recent review supports the first, but not the second conclusion.
Bourdillon et al pointed to the need to take into account age, gender and ethnicity.3 We identified five suitable studies in systematic review in the literature, which compared the sensitivity and specificity of the ECG (using the Sokolow-Lyon and Cornell criteria) for detecting LVH in African origin and White population groups.4-8
Specificity was high using two sets of criteria in the White groups (Cornell: 87.4%; Sokolow-Lyon: 88.9%) but was much lower in the African origin group using the Sokolow-Lyon criteria (72.1%). Specificity was higher for the African origin group using the Cornell criteria (86.2%). There was also some evidence that Cornell criteria were better for African origin populations than Sokolow-Lyon for sensitivity.
Our evidence favours the Cornell criteria over the Sokolow-Lyon criteria. While we agree with Pewsner et al that the ECG is not sufficient for diagnosing LVH we emphasise that it does not have equivalent validity across ethnic groups, which is important in our modern, multi-ethnic and globalised society .
1. Pewsner D, J�ni P, Egger M, Battaglia M, Sundstr�m J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. doi:10.1136/bmj.39276.636354.AE2.
2. Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limb leads. Am Heart J 1949;37:161-863.
3. Peter J Bourdillon. QRS voltage criteria for LVH can be useful. Response to Pewsner D, J�ni P, Egger M, Battaglia M, Sundstr�m J, Bachmann LM. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. doi:10.1136/bmj.39276.636354.AE
4. Okin PM, Wright JT, Nieminen MS, Jern S, Taylor AL, Phillips R, et al. Ethnic differences in electrocardiographic criteria for left ventricular hypertrophy: the LIFE study. American Journal of Hypertension 2002;15(8):663-671.
5. Chapman JN, Mayet J, Chang CL, Foale RA, Thom SAM, Poulter NR. Ethnic differences in the identification of left ventricular hypertrophy in the hypertensive patient. American Journal of Hypertension 1999;12(5):437.
6. Rautaharju PM MT, Siscovick D, Zhou SH, Gardin JM, Kronmal R, Furberg CD, Borhani NO, Newman AB. The CHS Collaborative Research Group. Classification accuracy of electrocardiographic criteria for left ventricular hypertrophy in normal weight and overweight older adults. Annals of Noninvasive Electrocardiology 1996;1:121-132.
7. Crow RS, Prineas RJ, Rautaharju P, Hannan P, Liebson PR. Relation Between Electrocardiography and Echocardiography for Left Ventricular Mass in Mild Systemic Hypertension (Results from Treatment of Mild Hypertension study). The American Journal of Cardiology 1995;75(17):1233.
8. Lee DK, Marantz PR, Devereux RB, Kligfield P, Alderman MH. Left ventricular hypertrophy in black and white hypertensives. Standard electrocardiographic criteria overestimate racial differences in prevalence. JAMA 1992;267(24):3294-3299.
Competing interests: None declared

Left ventricular hypertrophy and QT dispersion in hypertensive patients 11 October 2007

Antoni Sis� Almirall,
Associate Professor of Medicine. University of Barcelona
Les Corts and G�tic Primary Care Centers. Barcelona. Spain,
Antoni Dalf� Baqu�.
Send response to journal:
Re: Left ventricular hypertrophy and QT dispersion in hypertensive patients

Electrocardiography (ECG) is the usual method for identifying left ventricular hypertrophy (LVH) on standard hypertensive evaluation in primary care. Usually, diagnostic criteria for LVH by 12-lead ECG use a recognized criteria such as Sokolow-Lyon, Casale/Devereux, Cornell product, Cornell voltage, or a combination of variables such as QRS voltages, QRS duration and time-voltage QRS area (1). QT dispersion (QTd), as a measure of interlead variations of QT interval duration in the ECG, may serve as a measure of variability in ventricular recovery time. QT dispersion (QTd) is a recognised predictor of sudden death in patients with hypertrophic cardiomyopathy and congestive heart failure (2). The incidence of sudden death is also increased in hypertensive population with LVH (3).
We investigated the relationship between increased left ventricular mass index (LVMI) in hypertensive patients and QT dispersion of standard electrocardiograms, and calculated the sensitivity and specificity of a QTd value for determine LVH. We analized 125 patients (41 males and 84 females) with essential hypertension. Patients with bundle branch block, coronary heart disease or patients that had receiving an antiarrhythmic drugs were excluded from the study. Standard 12-leaded electrocardiograms were measured in sinus rhythm in all 12 leads by a single blinded observer. QT intervals measurements were obtained from a image-digitalized system (Snap-Scan AGFA). QT interval was taken from the beginning of QRS complex to the end of the T wave (return to isoelectric baseline). If U wave was present, QT interval was measured to the nadir of the curve between T and U waves. QTd was defined as the difference between the maximum (QTmax) and the minimum QT interval. A rate-corrected QT dispersion (QTdc) was calculated by Bazett's formula. Echocardiografic examination was evaluated in all patients according to the recommendation of the American Society of Echocardiography. M-mode echocardiographic left ventricular mass was analysed: criterion of LVH was LVMI above 134 gm/m2 in men and above 110 gm/m2 in women (4). Results are expressed as mean (SD) and statistical analysis was done by using the SPSS software package. Linear regression analysis was performed between all measured variables. QTd accurancy was analized using receiver operating characteristic (ROC) curves analysis.
There were 78 patients (62.4%) with LVH. The mean age were 64.3 (9.6 years), range 29 to 82. The Pearson�s correlation coefficient (r) between QTd and LVMI, QTdc and LVMI and QTmax and LVMI were 0.54 (p < 0.0001), 0.50 (p < 0.0001) and 0.34 (p < 0.0001), respectively. Test performance of QTd for identification of LVH was better in women than in men, with higher areas under the ROC curves in women (0.82 in women and 0.80 in men). In men, a QTd of 48.0 miliseconds provided a sensitivity of 89.5% (66.9-98.7) and specificity of 59.1% (36.4-79.3); otherwise, in women a QTd of 43.6 miliseconds provided a sensitivity of 83.05% (71-91.6) and a specificity of 64% (42.5-82).
These findings suggest that QTd is a useful noninvasive parameter for LVH detection in hypertensive patients, and may play an important role for improve identification of LVH by 12- lead ECG compared with other electrocardiographic criterias.
References.
1. Okin PM, Roman MJ, Devereux RB, Pickering TG, Borer JS, Kligfield P. Time-voltage QRS Area of the 12-lead electrocardiogram. Detection of left ventricular hypertrophy. Hypertension 1998;31:937-942.
2. Barr CS, Naas A, Freeman M, Lang CC, Struthers AD. QT dispersion and sudden unexpected death in chronic heart failure. The Lancet 1994; 343:327-9.
3. Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Int Med 1991; 114:345-52.
4. Devereux RB, Reicheck N. Echocardiographic determination of left ventricular mass in man. Anatomic validation of a method. Circulation 1977; 55:613-18.
Competing interests: None declared