- Email: [email protected]
Assessment of electrocardiographic criteria for left atrial enlargement in childhood
Assessment of ElectrocardiographicCriteria for Left Atrial Enlargement in Childhood JEFFREY MAOK, MD, and EHUD KRONGRAD, MD
Ninety children, aged 1 day to 18 years (median 7 months), with electrocardiographic or echocardiographic evidence of left atrial (LA) enlargement were selected to determine if electrocardiographic criteria accurately reflected increased LA dimension as determined by echocardiography. Four cardiac defects known to produce LA enlargement were chosen: ventricular septal defect (24 patients), patent ductus arteriosus (25 patients), cardiomyopathy (27 patients) and mitral regurgitation (14 patients). Different electrocardiographic criteria for LA enlargement were assessed. The data indicated that the overall sensitivity and predictive value of the ECG to detect LA enlargement were 40 and 85 %,
respectively. The ECG .and echocardiogram failed to agree in 62% of the patients. The most predictive variable for LA enlargement was the presence of a notched P wave in the limb leads with a large negative terminal deflection in lead VI. The sensitivity of ECG was highest in patients with chronic LA overload status, in mitral regurgitation (77 % )T cardiomyopathy (50 % ) and ventricular septal defect (54% ). The results show that in the pediatric population, electrocardiographic criteria are moderately predictive for LA enlargement but not as sensitive as generally believed.
Detection of left atria1 (LA) enlargement in pediatric patients is helpful in the diagnosis and assessment of severity of congenital cardiac defects with LA volume overload or acquired cardiac defects such as mitral valve pathology and cardiomyopathy. Detection of LA enlargement by electrocardiography has been studied in adult patients and has varied from 44 to 92%-s Data for the normal range of P-wave height and duration in lead II have been established for childreng,lO; but the only previous published study testing electrocardiographic criteria of LA enlargement in pediatric patients is the report by Reynolds’l assessing the “terminal P-VI index” as a measure of LA enlargement against radiologic evidence of LA enlargement.” This study determines the sensitivity and predictive value of electrocardiographic criteria for LA enlargement in pediatric patients using echocardiographically derived LA size.
Methods
(Am J Cardiol 1984;53:215-217)
Ninety patients aged 1 day to 18 years (median 7 months) were selected for this study because they fulfilled echocardiographic or electrocardiographic criteria for LA enlargement. Four cardiac defects known to cause enlargement of the left atrium were chosen for study: patent ductus arteriosus (25 patients), ventricular septal defect (24 patients), cardiomyopathy of different kinds (27 patients), and mitral regurgitation (14 patients). M-mode echocardiograms were obtained from the precordial window using standard techniques of positioning and angulation. LA dimension was measured in end-systole using a previously described technique and table of measurements based on the patient’s age, weight and height,” (Fig. 1). It has been further suggested that detection of LA enlargement may be enhanced by measurement of the LA/aortic ratio.‘:{ ‘5 Twelve-lead ECGs were recorded on a Hewlett-Packard model 1514C recorder using standard techniques of lead placement. Two criteria were selected for use in our study. The first criterion was a broad notched P wave in any limb lead with a duration >0.08 second.” The second criterion: a variant of the Morris index,16 was a negative terminal deflect,ion in lead VI that was equal to or exceeded the subsequent PK segment. The 2 criteria used are illustrated in Figure 2. If a patient satisfied criterion 1 or 2, LA enlargement was diagnosed. Statistical significance of difference was assessed by chisquare analysis. The sensitivity of a test, was defined as the
From the Division of Pediatric Cardiology, Department of Pediatrics, Babies Hospital and College of Physicians and Surgeons, Columbia University, New York, New York. Manuscript received January 24, 1983; revised manuscript received September 2, 1983, accepted September 10, 1983. Address for reprints: Ehud Krongrad, MD, Division of Pediatric Cardiology, Babies Hospital, 3959 Broadway, New York, New York
10032. 215
216
ELECTROCARDIOGRAPHICLEFT ATRIAL ENLARGEMENT
TABLE I
Sensitivity, Predictive Value and Failure of the Electrocardiogram and Echocardiogram to Agree for Each Specific Cardiac Lesion
Cardiovascular Lesion Cardiomyopathy Mitral insufficiency Patent ductus arteriosus Ventricular septal defect
Sensitivity 50 77 5:
Predictive Value
Fail to Agree 48
:: 100 47
:; 58
All values expressed as percentages.
Left Atriym
l
I
*
,‘r
ECG correctly identified 6 of 12 patients with an enlarged LA dimension or 10 of 12 patients with either an enlarged LA dimension or increased LA/aortic ratio. When criteria 1 and 2 were present on the ECG, the ECG and echocardiogram agreed in 16 of 18 patients (89%) and failed to agree in 2 of 18 patients (11%). Correlation of quantitative division of LA dimension with electrocardiographic evidence of LA enlargement (criterion 1 or 2) revealed that as LA dimension increased, the ECG became more sensitive. This was statistically significant. The sensitivity and predictive value of the ECG to detect LA enlargement in each specific cardiac lesion studied are given in Table I. Discussion
FIGURE 1. Echocardiogram of the left atrium and aortic root. The left atrial dimension is measured in end-systole between the arrows.
percentage of patients with abnormal findings who were correctly detected. The predictive value of an abnormal test was defined as the percentage of patients with abnormal findings who are truly abnormal; that is, sensitivity = true positive/(true positive + false negative) and predictive value = true positive/(true positive + false positive). Results The sensitivity and predictive value of the ECG to detect LA enlargement as assessed by LA dimension were 48 and 73%, respectively; as assessed by the LA/ aortic ratio they were 39 and 79%, respectively; and as assessed by either the LA dimension or LA/aortic ratio, they were 40 and 85%, respectively. The ECG and echocardiogram did not agree in 48% of the patients when assessed by LA dimension, 64% of the patients when assessed by the LA/aortic ratio and 62% of the patients when assessed by either the LA dimension or the LA/aortic ratio. Eighteen patients had criterion 1, a broad notched P wave on the ECG. Criterion 1 was most often noted in leads I and II, but occasionally occurred in leads I or III alone. Patients with criterion 1 tended to be older (median age 12 years) and clinically to have mitral regurgitation or cardiomyopathy. Twelve patients had criterion 2, a negative terminal deflection of the P wave in lead Vi that equalled or exceeded the subsequent PR segment. When present, the
We studied a group of patients with LA enlargement diagnosed by electrocardiographic and echocardiographic data and assessed the sensitivity and predictive value of the ECG. The overall sensitivity and predictive value of the ECG to detect LA enlargement were 40 and 85%, respectively. In assessing the sensitivity and predictive value of a test in detection of disease, a standard is required to ascertain the presence of the disease. In our study, we used LA dimension and LA/aortic root ratio as determined by M-mode echocardiography as the standard for assessing LA size, because previous studies have found a good correlation between LA size determined by angiography and echocardiography.17-lg The ECG had high positive predictive ability in patients with patent ductus arteriosus, mitral insufficiency and cardiomyopathy: 100,91 and 85%, respectively. In patients with ventricular septal defect, the ECG had low predictive ability to detect patients with LA enlargement (44%). The ECG was most sensitive for patients with mitral regurgitation (77%), moderately sensitive for patients with cardiomyopathy and ventricular septal defect (50 and 54%, respectively) and lowest for patients with patent ductus arteriosus (8%). This low sensitivity is probably secondary to the relatively short duration of LA dilatation in the patent ductus arteriosus as compared with the other 3 groups. The sensitivity of the ECG to detect LA enlargement may be lower than anticipated for several reasons. The left atrium is located posterior, leftward and superior to the right atrium and is depolarized after right atria1 depolarization. Therefore, LA depolarization contrib-
January 1. 1984
FIGURE
2. Electrocardiographic
criteria
1 and 2 for left atrial
enlargement.
utes to the middle and terminal thirds of the inscription of the P wave and it is only during the terminal third of the P wave that LA forces are unopposed.” Right atria1 forces make a greater contribution to the P wave than LA forces because of right atria1 proximity of the recording electrodes, that is, lead VI. Wave propagation and conduction velocity are slowed in the setting of volume and pressure overload.20 Volume and pressure loads on atrial tissue affect both structure and function.20-22 Therefore, the duration of volume and pressure overload may contribute to the appearance of LA enlargement on the ECG. Reynolds”” postulated that notching of the P wave occurred when the atria did not contract simultaneously, probably secondary to an intra- or interatrial conduction delay. The terminal portion of the P wave has been shown to represent electrical depolarization of the left atrium alone.” Josephson et a124studied patients with electrocardiographic evidence of LA enlargement by electrophysiologic techniques so that they could measure the high right atria1 to coronary sinus conduction time. They found that the patients with electrocardiographic evidence of LA enlargement tended to have prolonged interatrial conduction times, consistent with Reynolds’ hypothesis. Our study shows that although electrocardiographic criteria for detection of LA enlargement in pediatric patients are moderately predictive for patients with longstanding LA overload, the sensitivity was low. Furthermore, detection of LA enlargement by the ECG may be affected by different disease states, degree of severity and probably by the duration of the cardiac overload. References 1. Saunders JL, Calatayud JB, Schultz KJ, Maranaho V, Gooch AS, Goldberg H. Evaluation of electrocardiographic criteria for P wave abnormalities. Am Heart J 1967;74:757-765.
THE AMERICAN JOURNAL OF CARDIOLOGY
A, criterion
1, L,. B, criterion
Volume 53
217
2, lead V,
2. Rubler S, Shan NN, Moallem A. Comparison of left atrial size and pulmonary capillary pressure with P wave on the electrocardiogram. Am Heart J 1976;92:73-78. 3. Waggoner AD, Adyanthage AV, Quinnes MA, Alexander JK. Left atrial enlargement: echocardiographic assessment of electrocardiographic criteria. Circulation 1976;54:553-557. 4. Sololl LA, Zatuchni J. Relationships of the P wave to left atrial volume in rheumatic heart disease with mitral stenosis. Am J Med Sci 1958;235: 290-296. 5. Graybiel A, McFarland RA, Gates DC, Webster FA. Analyses of electrocardiograms obtained from 1000 young aviators. Am Heart J 1944;27: 524-549. 6. Green EW. Electrocardiographic pattern of atrial enlargement and abnormal impulse formation and conduction. In: Cassels DE, Ziegler RF, eds. Electrocardiography in Infants and Children. New York: Grune 8 Stratton, 1966:116-130. 7. Trounce JR. The electrocardiogram in mitral stenosis. Br Heart J 1952; 14:185-192. 8. Berliner K, Master AM. Mitral stenosis: correlation of electrocardiographic and pathologic observations. Arch Intern Med 1938:61:39-59. 9. Green EW. The normal P wave. In: Ref. 6:109-l 15. 10. Ziegler RF. Electrocardiographic Studies in Normal Infants and Children. Springfield, IL: Charles C Thomas, 195 1: 128- 129. 11. Reynolds JL. The electrocardiographic recognition of left atrial enlargement in childhood. Am Heart J 1967;74:179-191. 12. Meyer RA. Pediatric Echocardiography. Philadelphia: Lea & Febiger. 1977:292-293. 13. Brown OR, Harrison DC, Popp RL. An improved method for echocardiographic detection of left atrial enlargement. Circulation 1974;50:58-64. 14. Silverman NH, Lewis AB, Heymann MD, Rudolph AM. Echocardiographic assessment of ductus arteriosus shunt in premature infants. Circulation 1974;50:821-825. 15. Goldberg BB, Allen HD, Sahn DJ. Pediatric and Adolescent Echocardiography: A Handbook. Chicago: Year Book Medical, 1975:70. 16. Morris JJ, E&es EH, Whalen RE, Thompson HK, McIntosh HO. P wave analysis in valvular heart disease. Circulation 1964;29:242-252. 17. Yabek SM, Isabel JJ, Bhatt DR, Nakarawa M, Marks RA, Jarmakani JM. Echocardiographic determination of left atrial volumes in children with congenital heart disease. Circulation 1976;53:268-272. 18. Hirata T, Wolfe SB, Popp RL, Helman CH, Feigenbaum H. Estimation of left atrial size using ultrasound. Am Heart J 1969;78:43-52. 19. ten Cate FJ, Kloster RE, Van Dorp WG, Meuster GI, Roelandf J. Dimensions and volumes of left atrium and ventricle by single beam echocardiography. Br Heart J 1974;36:737-746. 20. Hordof AJ, Edie R, hlalm JR, Hoffman BF, Rosen MR. Eleclrophysiologic properties and response to pharmacologic agents of fibers from diseased human atria. Circulation 1976;54:774-779. 21. Pham TD, Wit AL, Hordof AJ, Malm JR, Fenoglio JJ. Right atrial ultrastructure in congenital heart disease. I. Comparison of ventricular septal defect and endocardial cushion defect. Am J Cardiol 1978;42:973-982. 22. Fenoglio JJ, Pham TD, Hordof AJ, Edie R, wit AL. Right atrial ultrastructure in congenital heart disease. II. Atrial septal defect: Effects of volume overload. Am J Cardiol 1979;43:820-827. 23. Reynolds G. The atrial electrogram in mitral stenosis. Br Heart J 1953; is~%in-3511 _.___ ___ 24. Josephson ME, Kastor JA, Monganroth J. Electrocardiographic left atrial enlargement: electrophysiologic, echocardiographic and hemodynamic correlates. Am J Cardiol 1977;39:967-971.
Ninety children, aged 1 day to 18 years (median 7 months), with electrocardiographic or echocardiographic evidence of left atrial (LA) enlargement were selected to determine if electrocardiographic criteria accurately reflected increased LA dimension as determined by echocardiography. Four cardiac defects known to produce LA enlargement were chosen: ventricular septal defect (24 patients), patent ductus arteriosus (25 patients), cardiomyopathy (27 patients) and mitral regurgitation (14 patients). Different electrocardiographic criteria for LA enlargement were assessed. The data indicated that the overall sensitivity and predictive value of the ECG to detect LA enlargement were 40 and 85 %,
respectively. The ECG .and echocardiogram failed to agree in 62% of the patients. The most predictive variable for LA enlargement was the presence of a notched P wave in the limb leads with a large negative terminal deflection in lead VI. The sensitivity of ECG was highest in patients with chronic LA overload status, in mitral regurgitation (77 % )T cardiomyopathy (50 % ) and ventricular septal defect (54% ). The results show that in the pediatric population, electrocardiographic criteria are moderately predictive for LA enlargement but not as sensitive as generally believed.
Detection of left atria1 (LA) enlargement in pediatric patients is helpful in the diagnosis and assessment of severity of congenital cardiac defects with LA volume overload or acquired cardiac defects such as mitral valve pathology and cardiomyopathy. Detection of LA enlargement by electrocardiography has been studied in adult patients and has varied from 44 to 92%-s Data for the normal range of P-wave height and duration in lead II have been established for childreng,lO; but the only previous published study testing electrocardiographic criteria of LA enlargement in pediatric patients is the report by Reynolds’l assessing the “terminal P-VI index” as a measure of LA enlargement against radiologic evidence of LA enlargement.” This study determines the sensitivity and predictive value of electrocardiographic criteria for LA enlargement in pediatric patients using echocardiographically derived LA size.
Methods
(Am J Cardiol 1984;53:215-217)
Ninety patients aged 1 day to 18 years (median 7 months) were selected for this study because they fulfilled echocardiographic or electrocardiographic criteria for LA enlargement. Four cardiac defects known to cause enlargement of the left atrium were chosen for study: patent ductus arteriosus (25 patients), ventricular septal defect (24 patients), cardiomyopathy of different kinds (27 patients), and mitral regurgitation (14 patients). M-mode echocardiograms were obtained from the precordial window using standard techniques of positioning and angulation. LA dimension was measured in end-systole using a previously described technique and table of measurements based on the patient’s age, weight and height,” (Fig. 1). It has been further suggested that detection of LA enlargement may be enhanced by measurement of the LA/aortic ratio.‘:{ ‘5 Twelve-lead ECGs were recorded on a Hewlett-Packard model 1514C recorder using standard techniques of lead placement. Two criteria were selected for use in our study. The first criterion was a broad notched P wave in any limb lead with a duration >0.08 second.” The second criterion: a variant of the Morris index,16 was a negative terminal deflect,ion in lead VI that was equal to or exceeded the subsequent PK segment. The 2 criteria used are illustrated in Figure 2. If a patient satisfied criterion 1 or 2, LA enlargement was diagnosed. Statistical significance of difference was assessed by chisquare analysis. The sensitivity of a test, was defined as the
From the Division of Pediatric Cardiology, Department of Pediatrics, Babies Hospital and College of Physicians and Surgeons, Columbia University, New York, New York. Manuscript received January 24, 1983; revised manuscript received September 2, 1983, accepted September 10, 1983. Address for reprints: Ehud Krongrad, MD, Division of Pediatric Cardiology, Babies Hospital, 3959 Broadway, New York, New York
10032. 215
216
ELECTROCARDIOGRAPHICLEFT ATRIAL ENLARGEMENT
TABLE I
Sensitivity, Predictive Value and Failure of the Electrocardiogram and Echocardiogram to Agree for Each Specific Cardiac Lesion
Cardiovascular Lesion Cardiomyopathy Mitral insufficiency Patent ductus arteriosus Ventricular septal defect
Sensitivity 50 77 5:
Predictive Value
Fail to Agree 48
:: 100 47
:; 58
All values expressed as percentages.
Left Atriym
l
I
*
,‘r
ECG correctly identified 6 of 12 patients with an enlarged LA dimension or 10 of 12 patients with either an enlarged LA dimension or increased LA/aortic ratio. When criteria 1 and 2 were present on the ECG, the ECG and echocardiogram agreed in 16 of 18 patients (89%) and failed to agree in 2 of 18 patients (11%). Correlation of quantitative division of LA dimension with electrocardiographic evidence of LA enlargement (criterion 1 or 2) revealed that as LA dimension increased, the ECG became more sensitive. This was statistically significant. The sensitivity and predictive value of the ECG to detect LA enlargement in each specific cardiac lesion studied are given in Table I. Discussion
FIGURE 1. Echocardiogram of the left atrium and aortic root. The left atrial dimension is measured in end-systole between the arrows.
percentage of patients with abnormal findings who were correctly detected. The predictive value of an abnormal test was defined as the percentage of patients with abnormal findings who are truly abnormal; that is, sensitivity = true positive/(true positive + false negative) and predictive value = true positive/(true positive + false positive). Results The sensitivity and predictive value of the ECG to detect LA enlargement as assessed by LA dimension were 48 and 73%, respectively; as assessed by the LA/ aortic ratio they were 39 and 79%, respectively; and as assessed by either the LA dimension or LA/aortic ratio, they were 40 and 85%, respectively. The ECG and echocardiogram did not agree in 48% of the patients when assessed by LA dimension, 64% of the patients when assessed by the LA/aortic ratio and 62% of the patients when assessed by either the LA dimension or the LA/aortic ratio. Eighteen patients had criterion 1, a broad notched P wave on the ECG. Criterion 1 was most often noted in leads I and II, but occasionally occurred in leads I or III alone. Patients with criterion 1 tended to be older (median age 12 years) and clinically to have mitral regurgitation or cardiomyopathy. Twelve patients had criterion 2, a negative terminal deflection of the P wave in lead Vi that equalled or exceeded the subsequent PR segment. When present, the
We studied a group of patients with LA enlargement diagnosed by electrocardiographic and echocardiographic data and assessed the sensitivity and predictive value of the ECG. The overall sensitivity and predictive value of the ECG to detect LA enlargement were 40 and 85%, respectively. In assessing the sensitivity and predictive value of a test in detection of disease, a standard is required to ascertain the presence of the disease. In our study, we used LA dimension and LA/aortic root ratio as determined by M-mode echocardiography as the standard for assessing LA size, because previous studies have found a good correlation between LA size determined by angiography and echocardiography.17-lg The ECG had high positive predictive ability in patients with patent ductus arteriosus, mitral insufficiency and cardiomyopathy: 100,91 and 85%, respectively. In patients with ventricular septal defect, the ECG had low predictive ability to detect patients with LA enlargement (44%). The ECG was most sensitive for patients with mitral regurgitation (77%), moderately sensitive for patients with cardiomyopathy and ventricular septal defect (50 and 54%, respectively) and lowest for patients with patent ductus arteriosus (8%). This low sensitivity is probably secondary to the relatively short duration of LA dilatation in the patent ductus arteriosus as compared with the other 3 groups. The sensitivity of the ECG to detect LA enlargement may be lower than anticipated for several reasons. The left atrium is located posterior, leftward and superior to the right atrium and is depolarized after right atria1 depolarization. Therefore, LA depolarization contrib-
January 1. 1984
FIGURE
2. Electrocardiographic
criteria
1 and 2 for left atrial
enlargement.
utes to the middle and terminal thirds of the inscription of the P wave and it is only during the terminal third of the P wave that LA forces are unopposed.” Right atria1 forces make a greater contribution to the P wave than LA forces because of right atria1 proximity of the recording electrodes, that is, lead VI. Wave propagation and conduction velocity are slowed in the setting of volume and pressure overload.20 Volume and pressure loads on atrial tissue affect both structure and function.20-22 Therefore, the duration of volume and pressure overload may contribute to the appearance of LA enlargement on the ECG. Reynolds”” postulated that notching of the P wave occurred when the atria did not contract simultaneously, probably secondary to an intra- or interatrial conduction delay. The terminal portion of the P wave has been shown to represent electrical depolarization of the left atrium alone.” Josephson et a124studied patients with electrocardiographic evidence of LA enlargement by electrophysiologic techniques so that they could measure the high right atria1 to coronary sinus conduction time. They found that the patients with electrocardiographic evidence of LA enlargement tended to have prolonged interatrial conduction times, consistent with Reynolds’ hypothesis. Our study shows that although electrocardiographic criteria for detection of LA enlargement in pediatric patients are moderately predictive for patients with longstanding LA overload, the sensitivity was low. Furthermore, detection of LA enlargement by the ECG may be affected by different disease states, degree of severity and probably by the duration of the cardiac overload. References 1. Saunders JL, Calatayud JB, Schultz KJ, Maranaho V, Gooch AS, Goldberg H. Evaluation of electrocardiographic criteria for P wave abnormalities. Am Heart J 1967;74:757-765.
THE AMERICAN JOURNAL OF CARDIOLOGY
A, criterion
1, L,. B, criterion
Volume 53
217
2, lead V,
2. Rubler S, Shan NN, Moallem A. Comparison of left atrial size and pulmonary capillary pressure with P wave on the electrocardiogram. Am Heart J 1976;92:73-78. 3. Waggoner AD, Adyanthage AV, Quinnes MA, Alexander JK. Left atrial enlargement: echocardiographic assessment of electrocardiographic criteria. Circulation 1976;54:553-557. 4. Sololl LA, Zatuchni J. Relationships of the P wave to left atrial volume in rheumatic heart disease with mitral stenosis. Am J Med Sci 1958;235: 290-296. 5. Graybiel A, McFarland RA, Gates DC, Webster FA. Analyses of electrocardiograms obtained from 1000 young aviators. Am Heart J 1944;27: 524-549. 6. Green EW. Electrocardiographic pattern of atrial enlargement and abnormal impulse formation and conduction. In: Cassels DE, Ziegler RF, eds. Electrocardiography in Infants and Children. New York: Grune 8 Stratton, 1966:116-130. 7. Trounce JR. The electrocardiogram in mitral stenosis. Br Heart J 1952; 14:185-192. 8. Berliner K, Master AM. Mitral stenosis: correlation of electrocardiographic and pathologic observations. Arch Intern Med 1938:61:39-59. 9. Green EW. The normal P wave. In: Ref. 6:109-l 15. 10. Ziegler RF. Electrocardiographic Studies in Normal Infants and Children. Springfield, IL: Charles C Thomas, 195 1: 128- 129. 11. Reynolds JL. The electrocardiographic recognition of left atrial enlargement in childhood. Am Heart J 1967;74:179-191. 12. Meyer RA. Pediatric Echocardiography. Philadelphia: Lea & Febiger. 1977:292-293. 13. Brown OR, Harrison DC, Popp RL. An improved method for echocardiographic detection of left atrial enlargement. Circulation 1974;50:58-64. 14. Silverman NH, Lewis AB, Heymann MD, Rudolph AM. Echocardiographic assessment of ductus arteriosus shunt in premature infants. Circulation 1974;50:821-825. 15. Goldberg BB, Allen HD, Sahn DJ. Pediatric and Adolescent Echocardiography: A Handbook. Chicago: Year Book Medical, 1975:70. 16. Morris JJ, E&es EH, Whalen RE, Thompson HK, McIntosh HO. P wave analysis in valvular heart disease. Circulation 1964;29:242-252. 17. Yabek SM, Isabel JJ, Bhatt DR, Nakarawa M, Marks RA, Jarmakani JM. Echocardiographic determination of left atrial volumes in children with congenital heart disease. Circulation 1976;53:268-272. 18. Hirata T, Wolfe SB, Popp RL, Helman CH, Feigenbaum H. Estimation of left atrial size using ultrasound. Am Heart J 1969;78:43-52. 19. ten Cate FJ, Kloster RE, Van Dorp WG, Meuster GI, Roelandf J. Dimensions and volumes of left atrium and ventricle by single beam echocardiography. Br Heart J 1974;36:737-746. 20. Hordof AJ, Edie R, hlalm JR, Hoffman BF, Rosen MR. Eleclrophysiologic properties and response to pharmacologic agents of fibers from diseased human atria. Circulation 1976;54:774-779. 21. Pham TD, Wit AL, Hordof AJ, Malm JR, Fenoglio JJ. Right atrial ultrastructure in congenital heart disease. I. Comparison of ventricular septal defect and endocardial cushion defect. Am J Cardiol 1978;42:973-982. 22. Fenoglio JJ, Pham TD, Hordof AJ, Edie R, wit AL. Right atrial ultrastructure in congenital heart disease. II. Atrial septal defect: Effects of volume overload. Am J Cardiol 1979;43:820-827. 23. Reynolds G. The atrial electrogram in mitral stenosis. Br Heart J 1953; is~%in-3511 _.___ ___ 24. Josephson ME, Kastor JA, Monganroth J. Electrocardiographic left atrial enlargement: electrophysiologic, echocardiographic and hemodynamic correlates. Am J Cardiol 1977;39:967-971.