LETTerS TO T"E EOITOR
....
definitions reviewed in this letter have already been reported?, 10 Robert S. Galen, MD, MPH Clinical Chemistry Director Francis Deiafield Hospital Division College of Physicians & Surgeons of Columbia University New York, New York References 1. Schweitzer P, Je|inek VM, Herman MV, el ah Comparison of the two*step and maximal exercise tests in patients with coronary artery disease. Am J Cardiol 33: 797-800, t974 2. Vecchio TJ: Predictive value of a single diagnostic test in unseiacted populations. N Engl J Med 274:1171-1 t73, 1966 3. Sunderman FW, Van Soestbergen AA: Probability computations for clinical interpretations of screening tests. Am J Clin Patho155:105-111, 1971 4. Sunderman FW: Conceptual problems in the interpretation of multitest surveys. In, Clinically Oriented Interpretation of Laboratory Data (Gabriali ER, ed). New York, Academic Press, 1972, p 39-68 5. Werner M, Brooks SH, Wette R: Strategy of cost-effective laboratory testing. Human Pathot 4:17-30, 1973 6. Holland WW, Whitehead TP: Value of new laboratory tests in diagnosis and treatment. Lancet 2:39t-394, 1974 7. Galen RS, Sebag J, Gambino SR: From data to information: the predictive value of laboratory tests as defined by variable referent values. Proc IRIA Symposium on Medical Data Processing, Toulouse, March 1974, p 183-190 8. Galen RS, Gambino SR: Beyond Normality--The Predictive Value of Medical Diagnosis. New York, John Wiley, 1975, in press 9. JelUffe RW: Quantitative aspects of clinical judgment. Am J Med 55:431-433, 1973 10. Schwartz WB, Gorry GA, Kassirer JP, et al: Decision analysis and clinical judgment. Am J Med 55:459-472, 1973
REPLY
In reply to Dr. Silverstein's scolding about not being scientific enough in m y comments about the significance of J point depression in the exercise electrocardiogram, the article was not m e a n t to be a rigorously designed scientific treatise, b u t rather an essay in which I sought to share experience with the reader. I tried to point out simply t h a t the appearance of J point depression cannot always be taken as an innocent finding. At times it is the only detectable electrocardiographic change during true anginal pain. I readily accept t h a t this change is frequently encountered in normal, healthy and well conditioned persons. I don't know how to distinguish the innocent J p o i n t depression from the bad except to look for corroborative evidence of myocardial ischemia such as chest pain or abnormal curunary angiographic findings. I indicated t h a t we are trying to devise a test to see if administration of nitroglycerin will selectively reverse the J point depression of true ischemia and perhaps help to differentiate the innocent from the bad. We, too, would like to know how to be sure about this point. Albert A. Kattus, MD Professor of Medicine The Center for the Health Sciences University of California Los Angeles, California
J POINT DEPRESSION IN THE EXERCISE E L E C T R O C A R D I O G R A M
T h e correlative method of presentation in the article by K a t t u s 1 surely must be viewed as excellent teaching. Those of us in exercise physiology are indebted to Dr. Kattus for his leadership in this field. However, I m u s t quibble with his lack of true scientific approach to the problem of J point depression, which he stated is "often a stage in the development of the ischemic response." This is a classic exa m p l e of the "post-hoc" fallacy. I would have appreciated his stating t h a t this is his bias on what may turn out to be a clinical fact but is not yet firmly established. However, his s t a t e m e n t on the significance of J point depression with no qualification other than "often" now exists in the literature. While this kind of prejudicial statement could be entirely true, the statistically significant evidence was not quoted and is lacking as far as I am aware. 2 K a t t u s ' statement will be quoted henceforth and without the "often." Patients wilt be found to have coronary disease on the basis of a normal upsloping J point depression when indeed they have no coronary disease at all. Figures of such stature as K a t t u s m u s t be careful to make clear when they are moving from verified scientific d a t a into the realm of impressions. Obviously it is my bias t h a t J point depression m a y be and usually is entirely normal and t h a t as yet there are no well accepted statistically valid criteria to differentiate normal from abnormal J point depression. I would appreciate anything Dr. Kattus has to offer concerning these criteria as well as clarification of his statement. H. Robert Silverstein, MD Division of Cardiology St. Francis Hosp!tal Hartford, Connecticut References 1. Kattus AA: Exorcise electrocardiography: recognition of the ischemic response, false positive and negative patterns. Am J Cardio133:72t-73t, 1974 2. Ellestadt MH, Allen W, Wan CK, et el.: Maximal treadmill stress test for cardiovascular evaluatlon. Circulation 39:517-522, 1969
538
October 6, 1975
The American Journal of CARDIOLOGY
EARLY HOSPITAL DISCHARGE AFTER M Y O C A R D I A L INFARCTION
In their study of early mobilization after myocardial infarction Bloch et al. 1 referred to a paper by us 2 pertaining to e a r l y discharge after acute myocardial infarction: "However, since only 15 percent of the patients hospitalized for infarction were a d m i t t e d to their study, the applicability of their conclusions is severely limited." We should like to take exception to this statement and make the following comments. Although it is true t h a t only 15 percent of the 925 consecutive patients with definite myocardial infarction screened by us were a d m i t t e d to the study, the absolute number of patients in both series is similar (138 in ours, 154 in theirs). The reasons for exclusion in our study should be examined more carefully. Twelve percent died before assessment for randomization; an additional 18 percent with p u m p failure on the 5th hospital day and 8 percent with heart block reflect the somewhat skewed nature of patients with myocardial infarction referred to our institution because of our particular interest in the t r e a t m e n t of p u m p failure. A further 8 percent were excluded because of complicating medical illness necessitating a longer stay, and 7 percent were rejected for "social" reasons. Thus there is good reason to believe t h a t m a n y more patients would be eligible in most hospitals where more "typical" cases of myocardial infarction are seen. We studied the problem of early discharge as well as early mobilization by randomly assigning our patients to two groups, one discharged at 2 weeks and the other at 3 weeks, and found no difference in subsequent outcome between the two groups. The average length of stay of the early mobilization group of Bloch et al., 21.3 days, ~orresponds to t h a t of the "long stay" group in our study. The point is t h a t most patients with uncomplicated acute myocardial infarction are candidates not only for early mobilization, b u t also for early discharge. Most can be
Volume 36
LETTERS TO THE EDITOR
discharged a t 2 weeks, and there is little justification for keeping t h e m longer t h a n 3 weeks. The importance of both our paper and t h a t of Bloch et al. is that, despite the fact t h a t several studies of early mobilization and discharge have been reported, these are the only two t h a t are truly randomized.
strated to accomplish considerable benefits without excessive risks. Antoine Bloch, MD Hbpital Cantonal Geneva, Switzerland Massachusetts General Hospital Boston, Massachusetts Henry Blackburn, MD, FACC University of Minnesota Minneapolis, Minnesota
Roman W. DeSanctis, MD, FACC Adolph M. Hutter, Jr., MD, FACC Massachusetts General Hospital Boston, Massachusetts References 1. Sloch A, Maede J-P, Haissly J-C, et ah Early mobilization after myocardial infarction. A controlled study. Am J Cardio134:152-157, 1974 2. Huller AM Jr, Sldel VW, Shine KI, el ah Early hospital discharge after myocardial infarction. N Engl J Med 288:1141-1143, 1973
REPLY
The study by H u t t e r et al. is an i m p o r t a n t randomized study. Our comment about their report was directed at the low proportion of all cases with acute myocardial infarctiofi included (15 vs. 75 percent in our study). The mortality rate before randomization was about the same in the two Series. In our study, patients Were not excluded because of medical illness necessitating a longer stay, thus lengthening somewhat the average hospital stay of the early mobilization group. The principal difference in the material of the two studies is the percentage of patients excluded because of complicated myocardial infarction (50 percent in the H u t t e r study vs. 15 percent in our study). This difference is partly related to the skewed nature of patients referred to the cardiac unit of the Massachusetts General Hospital, and partly due to the fact t h a t the Geneva definition of uncomplicated infarction was far less restrictive. In addition, in H u t t e r ' s study patients were randomly assigned into groups on the 6th hospital day, after exclusion of patients with ventricular fibrillation, ventricular tachycardia, asystole, second degree heart block or complete heart block; persistence of more than five ventricular premature beats per minute, multifocal ventricular premature beats or paired ventricular premature beats into the 5th day; continued requirement of any antiarrhythmic t h e r a p y into the 5th day; persistence of any evidence of congestive h e a r t failure into the 5th day; any hypotension requiring vasopressor agents; definite thromboembolic complications during the first 5 days; coronary pain persisting into the 5th day; occurrence of any acute episodes of infarction within the preceding 6 months. In contrast, in our study patients were randomly assigned into groups 24 to 48 hours after admission, and the only causes of exclusion were severe heart failure, cardiogenic shock, severe r h y t h m disturbances, severe thoracic pain or severe psychological problems. The study of H u t t e r et al. showed the benefit of early mobilization and early discharge after uncomplicated myocardial infarction. We d e m o n s t r a t e d t h a t this practice can be applied to the majority of patients a d m i t t e d for myocardial infarction in a nonreferral general hospital. Finally, we suspect t h a t both of our reports overemphasized somewhat the absence of statistically significant differences. However, due to relatively small sample size and number of events, the statistical power was not sufficient to establish t h a t there were absolutely no differences. But the thrust of these two truly randomized studies is t h a t wellplanned early mobilization is now unequivocally demon-
METHYLPREDNISOLONE IN ACUTE MYOCARDIAL INFARCTION
T h e study by Vyden et al. 1 requires some clarification. It appears from the experimental protocol t h a t the authors are treating not simple acute myocardial infarction, but cardiogenic shock (systolic blood pressure below 80 m m Hg, sustained for 30 minutes). 2 In their preparation, mean systemic vascular resistance was increased concomitantly with the drop in cardiac index. In this condition, it has long been known t h a t glucocorticosteroids reduce peripheral vasoconstriction, thereby increasing the stroke volume of the left ventricle. 3,4 In myocardial infarction without shock, methylprednisolone has not produced any hemodynamic changes. 5 Thus, t h e beneficial effect of glucocorticosteroids in acute myocardial infarction must be advocated on a basis other than hemodynamic, and the beneficial hemodynamic effects seen in these experiments probably cannot be extrapolated to the human situation for the reasons the authors themselves enumerate. Mario Feola, MD, FACC Department of Surgery Jefferson Medical College Philadelphia, Pennsylvania References 1. Vyden JK, Nagasawa K, Rabinowitz B, et al: Effects of methylprednisoloneadministration in acute myocardial infarction. Am J Cardiot 34:677-686, 1974 2. Agress CM, Rosenberg MJ, Jacobs HI, at al: Protracted shock in the closed-chest
dog following coronary embolization with graded microspheres. Am J Physiol 170: 536, 1952 3. Sambhi MP, Weil MH, Udhoji UN: Acute pharmacodynamic effect of glucocorticoids: cardiac output and related hemodynamic changes in normal subjects and patients in shock. Circulation 31:523, 1965 4, Diefzman RH, Ullehei RC: The treatment of cardiogenic shock. V. The use of corticosteroids in the treatment of cardiogenic shock. Am Heart J 75:274, 1968 5. Spath JA, Lane DL, Lefer AM: Protective action of methylprednisolone on the myocardium during experimental myocardial ischemia in the cat. Circ Res 35:44, 1974
REPLY
Feola states t h a t "mean systemic vascular resistance was increased concomitantly with the drop in cardiac index" when acute myocardial infarction was produced in our preparation and draws certain conclusions from this. However, mean systemic vascular resistance was not significantly increased with the production of infarction, changing from 5.71 • 0.46 to 6.88 ± 0.76 (t = 1.41), and thus the conclusions drawn are not valid. Most clinicians treating acute myocardial infarction would agree cardiogenic shock is accompanied by a marked fall in stroke work index coupled with a significant increase to abnormal levels in left ventricular end-diastolic pressure. However, in our preparation there was no change in left ventricular end-diastolic pressure of any significance
October 6, 1975
The American Journal of CARDIOLOGY
Volume 36
539
....
definitions reviewed in this letter have already been reported?, 10 Robert S. Galen, MD, MPH Clinical Chemistry Director Francis Deiafield Hospital Division College of Physicians & Surgeons of Columbia University New York, New York References 1. Schweitzer P, Je|inek VM, Herman MV, el ah Comparison of the two*step and maximal exercise tests in patients with coronary artery disease. Am J Cardiol 33: 797-800, t974 2. Vecchio TJ: Predictive value of a single diagnostic test in unseiacted populations. N Engl J Med 274:1171-1 t73, 1966 3. Sunderman FW, Van Soestbergen AA: Probability computations for clinical interpretations of screening tests. Am J Clin Patho155:105-111, 1971 4. Sunderman FW: Conceptual problems in the interpretation of multitest surveys. In, Clinically Oriented Interpretation of Laboratory Data (Gabriali ER, ed). New York, Academic Press, 1972, p 39-68 5. Werner M, Brooks SH, Wette R: Strategy of cost-effective laboratory testing. Human Pathot 4:17-30, 1973 6. Holland WW, Whitehead TP: Value of new laboratory tests in diagnosis and treatment. Lancet 2:39t-394, 1974 7. Galen RS, Sebag J, Gambino SR: From data to information: the predictive value of laboratory tests as defined by variable referent values. Proc IRIA Symposium on Medical Data Processing, Toulouse, March 1974, p 183-190 8. Galen RS, Gambino SR: Beyond Normality--The Predictive Value of Medical Diagnosis. New York, John Wiley, 1975, in press 9. JelUffe RW: Quantitative aspects of clinical judgment. Am J Med 55:431-433, 1973 10. Schwartz WB, Gorry GA, Kassirer JP, et al: Decision analysis and clinical judgment. Am J Med 55:459-472, 1973
REPLY
In reply to Dr. Silverstein's scolding about not being scientific enough in m y comments about the significance of J point depression in the exercise electrocardiogram, the article was not m e a n t to be a rigorously designed scientific treatise, b u t rather an essay in which I sought to share experience with the reader. I tried to point out simply t h a t the appearance of J point depression cannot always be taken as an innocent finding. At times it is the only detectable electrocardiographic change during true anginal pain. I readily accept t h a t this change is frequently encountered in normal, healthy and well conditioned persons. I don't know how to distinguish the innocent J p o i n t depression from the bad except to look for corroborative evidence of myocardial ischemia such as chest pain or abnormal curunary angiographic findings. I indicated t h a t we are trying to devise a test to see if administration of nitroglycerin will selectively reverse the J point depression of true ischemia and perhaps help to differentiate the innocent from the bad. We, too, would like to know how to be sure about this point. Albert A. Kattus, MD Professor of Medicine The Center for the Health Sciences University of California Los Angeles, California
J POINT DEPRESSION IN THE EXERCISE E L E C T R O C A R D I O G R A M
T h e correlative method of presentation in the article by K a t t u s 1 surely must be viewed as excellent teaching. Those of us in exercise physiology are indebted to Dr. Kattus for his leadership in this field. However, I m u s t quibble with his lack of true scientific approach to the problem of J point depression, which he stated is "often a stage in the development of the ischemic response." This is a classic exa m p l e of the "post-hoc" fallacy. I would have appreciated his stating t h a t this is his bias on what may turn out to be a clinical fact but is not yet firmly established. However, his s t a t e m e n t on the significance of J point depression with no qualification other than "often" now exists in the literature. While this kind of prejudicial statement could be entirely true, the statistically significant evidence was not quoted and is lacking as far as I am aware. 2 K a t t u s ' statement will be quoted henceforth and without the "often." Patients wilt be found to have coronary disease on the basis of a normal upsloping J point depression when indeed they have no coronary disease at all. Figures of such stature as K a t t u s m u s t be careful to make clear when they are moving from verified scientific d a t a into the realm of impressions. Obviously it is my bias t h a t J point depression m a y be and usually is entirely normal and t h a t as yet there are no well accepted statistically valid criteria to differentiate normal from abnormal J point depression. I would appreciate anything Dr. Kattus has to offer concerning these criteria as well as clarification of his statement. H. Robert Silverstein, MD Division of Cardiology St. Francis Hosp!tal Hartford, Connecticut References 1. Kattus AA: Exorcise electrocardiography: recognition of the ischemic response, false positive and negative patterns. Am J Cardio133:72t-73t, 1974 2. Ellestadt MH, Allen W, Wan CK, et el.: Maximal treadmill stress test for cardiovascular evaluatlon. Circulation 39:517-522, 1969
538
October 6, 1975
The American Journal of CARDIOLOGY
EARLY HOSPITAL DISCHARGE AFTER M Y O C A R D I A L INFARCTION
In their study of early mobilization after myocardial infarction Bloch et al. 1 referred to a paper by us 2 pertaining to e a r l y discharge after acute myocardial infarction: "However, since only 15 percent of the patients hospitalized for infarction were a d m i t t e d to their study, the applicability of their conclusions is severely limited." We should like to take exception to this statement and make the following comments. Although it is true t h a t only 15 percent of the 925 consecutive patients with definite myocardial infarction screened by us were a d m i t t e d to the study, the absolute number of patients in both series is similar (138 in ours, 154 in theirs). The reasons for exclusion in our study should be examined more carefully. Twelve percent died before assessment for randomization; an additional 18 percent with p u m p failure on the 5th hospital day and 8 percent with heart block reflect the somewhat skewed nature of patients with myocardial infarction referred to our institution because of our particular interest in the t r e a t m e n t of p u m p failure. A further 8 percent were excluded because of complicating medical illness necessitating a longer stay, and 7 percent were rejected for "social" reasons. Thus there is good reason to believe t h a t m a n y more patients would be eligible in most hospitals where more "typical" cases of myocardial infarction are seen. We studied the problem of early discharge as well as early mobilization by randomly assigning our patients to two groups, one discharged at 2 weeks and the other at 3 weeks, and found no difference in subsequent outcome between the two groups. The average length of stay of the early mobilization group of Bloch et al., 21.3 days, ~orresponds to t h a t of the "long stay" group in our study. The point is t h a t most patients with uncomplicated acute myocardial infarction are candidates not only for early mobilization, b u t also for early discharge. Most can be
Volume 36
LETTERS TO THE EDITOR
discharged a t 2 weeks, and there is little justification for keeping t h e m longer t h a n 3 weeks. The importance of both our paper and t h a t of Bloch et al. is that, despite the fact t h a t several studies of early mobilization and discharge have been reported, these are the only two t h a t are truly randomized.
strated to accomplish considerable benefits without excessive risks. Antoine Bloch, MD Hbpital Cantonal Geneva, Switzerland Massachusetts General Hospital Boston, Massachusetts Henry Blackburn, MD, FACC University of Minnesota Minneapolis, Minnesota
Roman W. DeSanctis, MD, FACC Adolph M. Hutter, Jr., MD, FACC Massachusetts General Hospital Boston, Massachusetts References 1. Sloch A, Maede J-P, Haissly J-C, et ah Early mobilization after myocardial infarction. A controlled study. Am J Cardio134:152-157, 1974 2. Huller AM Jr, Sldel VW, Shine KI, el ah Early hospital discharge after myocardial infarction. N Engl J Med 288:1141-1143, 1973
REPLY
The study by H u t t e r et al. is an i m p o r t a n t randomized study. Our comment about their report was directed at the low proportion of all cases with acute myocardial infarctiofi included (15 vs. 75 percent in our study). The mortality rate before randomization was about the same in the two Series. In our study, patients Were not excluded because of medical illness necessitating a longer stay, thus lengthening somewhat the average hospital stay of the early mobilization group. The principal difference in the material of the two studies is the percentage of patients excluded because of complicated myocardial infarction (50 percent in the H u t t e r study vs. 15 percent in our study). This difference is partly related to the skewed nature of patients referred to the cardiac unit of the Massachusetts General Hospital, and partly due to the fact t h a t the Geneva definition of uncomplicated infarction was far less restrictive. In addition, in H u t t e r ' s study patients were randomly assigned into groups on the 6th hospital day, after exclusion of patients with ventricular fibrillation, ventricular tachycardia, asystole, second degree heart block or complete heart block; persistence of more than five ventricular premature beats per minute, multifocal ventricular premature beats or paired ventricular premature beats into the 5th day; continued requirement of any antiarrhythmic t h e r a p y into the 5th day; persistence of any evidence of congestive h e a r t failure into the 5th day; any hypotension requiring vasopressor agents; definite thromboembolic complications during the first 5 days; coronary pain persisting into the 5th day; occurrence of any acute episodes of infarction within the preceding 6 months. In contrast, in our study patients were randomly assigned into groups 24 to 48 hours after admission, and the only causes of exclusion were severe heart failure, cardiogenic shock, severe r h y t h m disturbances, severe thoracic pain or severe psychological problems. The study of H u t t e r et al. showed the benefit of early mobilization and early discharge after uncomplicated myocardial infarction. We d e m o n s t r a t e d t h a t this practice can be applied to the majority of patients a d m i t t e d for myocardial infarction in a nonreferral general hospital. Finally, we suspect t h a t both of our reports overemphasized somewhat the absence of statistically significant differences. However, due to relatively small sample size and number of events, the statistical power was not sufficient to establish t h a t there were absolutely no differences. But the thrust of these two truly randomized studies is t h a t wellplanned early mobilization is now unequivocally demon-
METHYLPREDNISOLONE IN ACUTE MYOCARDIAL INFARCTION
T h e study by Vyden et al. 1 requires some clarification. It appears from the experimental protocol t h a t the authors are treating not simple acute myocardial infarction, but cardiogenic shock (systolic blood pressure below 80 m m Hg, sustained for 30 minutes). 2 In their preparation, mean systemic vascular resistance was increased concomitantly with the drop in cardiac index. In this condition, it has long been known t h a t glucocorticosteroids reduce peripheral vasoconstriction, thereby increasing the stroke volume of the left ventricle. 3,4 In myocardial infarction without shock, methylprednisolone has not produced any hemodynamic changes. 5 Thus, t h e beneficial effect of glucocorticosteroids in acute myocardial infarction must be advocated on a basis other than hemodynamic, and the beneficial hemodynamic effects seen in these experiments probably cannot be extrapolated to the human situation for the reasons the authors themselves enumerate. Mario Feola, MD, FACC Department of Surgery Jefferson Medical College Philadelphia, Pennsylvania References 1. Vyden JK, Nagasawa K, Rabinowitz B, et al: Effects of methylprednisoloneadministration in acute myocardial infarction. Am J Cardiot 34:677-686, 1974 2. Agress CM, Rosenberg MJ, Jacobs HI, at al: Protracted shock in the closed-chest
dog following coronary embolization with graded microspheres. Am J Physiol 170: 536, 1952 3. Sambhi MP, Weil MH, Udhoji UN: Acute pharmacodynamic effect of glucocorticoids: cardiac output and related hemodynamic changes in normal subjects and patients in shock. Circulation 31:523, 1965 4, Diefzman RH, Ullehei RC: The treatment of cardiogenic shock. V. The use of corticosteroids in the treatment of cardiogenic shock. Am Heart J 75:274, 1968 5. Spath JA, Lane DL, Lefer AM: Protective action of methylprednisolone on the myocardium during experimental myocardial ischemia in the cat. Circ Res 35:44, 1974
REPLY
Feola states t h a t "mean systemic vascular resistance was increased concomitantly with the drop in cardiac index" when acute myocardial infarction was produced in our preparation and draws certain conclusions from this. However, mean systemic vascular resistance was not significantly increased with the production of infarction, changing from 5.71 • 0.46 to 6.88 ± 0.76 (t = 1.41), and thus the conclusions drawn are not valid. Most clinicians treating acute myocardial infarction would agree cardiogenic shock is accompanied by a marked fall in stroke work index coupled with a significant increase to abnormal levels in left ventricular end-diastolic pressure. However, in our preparation there was no change in left ventricular end-diastolic pressure of any significance
October 6, 1975
The American Journal of CARDIOLOGY
Volume 36
539
