The article by Schwartz and his colleagues (June 11 issue) (1) is the fifth such report by this group implicating prolongation of the QT interval in the sudden infant death syndrome (SIDS). However, independent confirmation of their conclusions is still lacking after 22 years. To the contrary, four prospective studies in addition to that by Southall et al. (2) (which was criticized by Schwartz and colleagues) have found no significant increase in the corrected QT interval (QTc) in infants who died of SIDS as compared with controls.

A major problem of the developmental long-QT interval theory of SIDS is that the mechanism of death should logically be a ventricular arrhythmia (specifically, torsade de pointes), if these infants are similar to patients with the genetic long-QT syndrome. Schwartz and colleagues present no evidence of ventricular arrhythmias in the 24 infants with a prolonged QTc who later died of SIDS, nor has anyone else recorded ventricular arrhythmias in otherwise normal infants. Yet, thousands of infants have been monitored electronically in the hospital and at home. In contrast, patients with clinically established long-QT syndrome have multiple episodes of ventricular arrhythmias, and death without prior arrhythmia was reported in only 9 percent in a study by Garson et al. (3)

These objections, plus the subjective nature of determining the QT interval and the lack of reproducibility of the results, (4) would make screening electrocardiography of all normal infants an extraordinary waste of medical resources, not to mention the fact that it would cruelly alarm thousands of parents about cardiac arrhythmias. Even if the screening were somehow successful, prevention could not be ensured, since Moss and Robinson, (5) in a study of the largest registry of patients with the long-QT syndrome in the world, found no statistically beneficial effect of any treatment for this condition.

Warren G. Guntheroth, M.D.
Philip S. Spiers, Ph.D.
University of Washington School of Medicine
Seattle, WA 98195

References

1. Schwartz PJ, Stramba-Badiale M, Segantini A, et al. Prolongation of the QT interval and the sudden infant death syndrome. N Engl J Med 1998;338:1709-14.
Return to Text

2. Southall DP, Arrowsmith WA, Stebbens V, Alexander JR. QT interval measurements before sudden infant death syndrome. Arch Dis Child 1986;61:327-33.
Return to Text

3. Garson A Jr, Dick M II, Fournier A, et al. The long QT syndrome in children: an international study of 287 patients. Circulation 1993;87:1866-72.
Return to Text

4. Towbin JA, Friedman RA. Prolongation of the QT and the sudden infant death syndrome. N Engl J Med 1998;338:1760-1.
Return to Text

5. Moss AJ, Robinson J. Clinical features of the idiopathic long QT syndrome. Circulation 1992;85:Suppl I:I-140-I-144.
Return to Text

Schwartz et al. correlate electrocardiographic abnormalities with subsequent death in infants. The main finding of their study is that infants who died of SIDS tended to have a much longer average QTc than those who survived (435 vs. 400 msec, P<0.01). What the authors neglect to discuss, however, is that those who died of SIDS also tended to have a much larger standard deviation in the QTc than those who survived (45 vs. 20 msec, P<0.001).

The greater observed variability in QTc in infants who died of SIDS merits further analysis. In particular, does this represent greater variation within each individual infant (QTc dispersion) or greater variation between infants (population diversity)? Settling this question should be easy, since the 12-lead electrocardiograms that were obtained for each infant are sufficient for identifying the degree of QTc dispersion in each infant. (1)

The identification of QTc dispersion could have clinical implications. First, this finding might represent a gain in sensitivity -- that is, it might be positive in some of the 12 infants who had a normal QTc yet died of SIDS. Second, this finding might represent a gain in specificity -- that is, it might not be mistakenly positive in 1 of 40 infants who survived. Third, this finding might corroborate other research on QTc dispersion as a means of predicting arrhythmias in adults. (2,3,4) These hypotheses could be tested quite quickly with use of the available data.

David R. Dancey, M.D.
Donald A. Redelmeier, M.D.
University of Toronto
Toronto, ON M4N 3M5, Canada

References

1. Day CP, McComb JM, Campbell RW. QT dispersion: an indication of arrhythmia risk in patients with long QT intervals. Br Heart J 1990;63:342-4.
Return to Text

2. Hii JT, Wyse DG, Gillis AM, Duff HJ, Solylo MA, Mitchell LB. Precordial QT interval dispersion as a marker of torsade de pointes: disparate effects of class IA antiarrhythmic drugs and amiodarone. Circulation 1992;86:1376-82.
Return to Text

3. Manttari M, Oikarinen L, Manninen V, Viitasolo M. QT dispersion as a risk factor for sudden cardiac death and fatal myocardial infarction in a coronary risk population. Heart 1997;78:268-72.
Return to Text

4. Goldner B, Brandspiegel HZ, Horwitz L, Jadonath R, Cohen TJ. Utility of QT dispersion combined with the signal-averaged electrocardiogram in detecting patients susceptible to ventricular tachyarrhythmia. Am J Cardiol 1995;76:1192-4.
Return to Text

The study by Schwartz et al. raises important methodologic and editorial questions. Although the study is described as prospective, it appears that the electrocardiograms were analyzed on a retrospective basis. It is not stated when the retrospective electrocardiographic analysis was performed and by whom, or whether all the observers were unaware of the clinical data on the subjects -- an essential prerequisite to avoid bias. Was the reproducibility of the measurements tested, given the subjective nature of measuring the QT interval?

Eric Rosenthal, M.D., M.R.C.P.
Guy's Hospital
London SE1 9RT, United Kingdom

The study by Schwartz et al. has clearly advanced our knowledge of cardiac culprits in SIDS. However, the engineer in me is troubled each time the unit milliseconds is used with the QTc. The teaching of Bazett's formula (QTc equals the QT interval in seconds divided by the square root of the RR interval in seconds) to students and residents would be enhanced by using the correct units (seconds^1/2) and by not moving the decimal point.

Thomas D. Scholz, M.D.
University of Iowa
Iowa City, IA 52242-1083

Schwartz et al. report compelling evidence of the links between SIDS and the long-QT syndrome. In their accompanying editorial, Towbin and Friedman (1) state that "half the infants who died of SIDS (12 of 24) and none of the survivors or the infants who died of other causes had a prolonged QTc." The original article states that "the 97.5th percentile for the QTc among the infants was 440 msec" and "the absolute risk of SIDS... of infants with a prolonged QTc was 1.53 percent." Both results mean that the great majority of neonates with a prolonged QTc will not be victims of SIDS.

Amaya Yoldi, M.D.
Francisca Sena, M.D.
Lluis Gutierrez, M.D.
Hospital Verge de la Cinta
Tortosa, Spain 43500

References

1. Towbin JA, Friedman RA. Prolongation of the QT interval and the sudden infant death syndrome. N Engl J Med 1998;338:1760-1.
Return to Text

To the Editor:

Drs. Guntheroth and Spiers mention four prospective studies that found no increase in the QTc of infants who died of SIDS but do not give specific citations. We are unaware of any other prospective study with adequate statistical power to demonstrate or disprove differences between groups, nor did we criticize the study by Southall et al. (1)

Why should infants who die of SIDS because of respiratory or other problems have a prolonged QT interval? A prolonged QT interval can be expected only in those who die of cardiac arrhythmia. Thus, one has to calculate whether the number of infants with a prolonged QT interval who died of SIDS exceeds that expected by chance alone according to the normal distribution of values for the QT interval. This is what happened in the study by Southall et al., as we discussed in our article.

That "independent confirmation of [our] conclusions is still lacking" is a consequence of the fact that no one else has carried out a study with numbers comparable to ours. Of course we did not provide data on arrhythmias in the infants with a prolonged QT interval who died of SIDS; we only recorded a standard electrocardiogram for 30 seconds. The fact that thousands of infants have been monitored without evidence of arrhythmias means nothing if they were not at risk for death from cardiac causes.

The statement by Guntheroth and Spiers that patients with the long-QT syndrome have multiple episodes of arrhythmia is irrelevant because it refers to children and adults. Our most recent data (2) indicate that the risk of death as a first cardiac event ranges between 2 and 20 percent according to the genotype and, a fact that is highly relevant here, is particularly high in the first year of life. (3) A child or adult who has an arrhythmic syncope loses consciousness and falls to the ground, whereas if an infant has a transient syncopal arrhythmia, where can he or she fall, since he or she is lying in a crib? A nonlethal episode could easily go unnoticed.

Finally, the data from the International Long-QT Syndrome Registry indicate a very high degree of efficacy of beta-blockers in reducing life-threatening arrhythmias in patients with the long-QT syndrome. (3,4) If an infant is recognized to be at high risk for lethal arrhythmias because of a markedly prolonged QT interval, then prophylactic therapy with beta-blockers for at least several months is a safe and rational approach.

Drs. Dancey and Redelmeier misrepresent our data by saying that the QTc of the infants who died of SIDS "tended" to be longer than that of the survivors. Since the values (435±45 vs. 400±20 msec) were significantly different (P<0.01), the QTc of the infants who died of SIDS actually was longer. We did not discuss the difference in the standard deviations of the QTc between the two groups because the larger standard deviation among the infants who died of SIDS is the consequence of the fact that some infants had a normal QTc and others had a markedly prolonged QTc. Thus, the greater standard deviation of the QTc represents variation in the group of infants and not in individual infants. Indeed, we did not observe significantly greater QT dispersion among the infants who died of SIDS than among the control infants who did not.

Dr. Rosenthal gives us a chance to clarify the important issue of blinding. The investigator measuring the QT interval was always unaware of the outcome of the infants. The study was indeed prospective, since we defined a priori the cutoff value for a prolonged QT interval as the traditional 440 msec, which also represented a value 2 SD above the mean.

Although the engineer in Dr. Scholz is troubled by the units in Bazett's formula, the biologists in us could not care less. What matters for the survival of patients is the duration of the QT interval in relation to the cycle length, and not the units in which it is expressed. The use of seconds or milliseconds provides the relevant biologic information; the presentation of these values as seconds^1/2 adds nothing. The vexing question of the "proper" unit to use with Bazett's formula has been answered by the careful analysis of Molnar et al. (5) They realized that when Taran and Szilagyi (6) re-expressed Bazett's formula in the form currently used, they eliminated one step (dividing the value by one second), which allowed the expression of the QTc in the unit that is both logical and arithmetically correct (namely, that of the original QT interval). Thus, we will gladly continue to express QTc in milliseconds, knowing what it means.

The comment by Yoldi et al. is essentially correct. Our statement that "none of the other infants had a prolonged QTc" obviously refers to the infants who died of causes other than SIDS. Since we specified in the text that 97.5 percent of the infants in the study had a QTc of 440 msec or less, it logically follows that 2.5 percent had a QTc exceeding this value.

Peter J. Schwartz, M.D.
Marco Stramba-Badiale, M.D., Ph.D.
University of Pavia
27100 Pavia, Italy

References

1. Southall DP, Arrowsmith WA, Stebbens V, Alexander JR. QT interval measurements before sudden infant death syndrome. Arch Dis Child 1986;61:327-33.
Return to Text

2. Zareba W, Moss AJ, Schwartz PJ, et al. Influence of the genotype on the clinical course of the long-QT syndrome. N Engl J Med 1998;339:960-5.
Return to Text

3. Schwartz PJ, Priori SG, Napolitano C. Long QT syndrome. In: Zipes DP, Jalife J, eds. Cardiac electrophysiology: from cell to bedside. 3rd ed. Philadelphia: W.B. Saunders (in press).
Return to Text

4. Moss AJ, Schwartz PJ, Crampton RS, Locati E, Carleen E. The long QT syndrome: a prospective international study. Circulation 1985;71:17-21.
Return to Text

5. Molnar J, Weiss JS, Rosenthal JE. The missing second: what is the correct unit for the Bazett corrected QT interval? Am J Cardiol 1995;75:537-8.
Return to Text

6. Taran LM, Szilagyi N. The duration of the electrical systole (Q-T) in acute rheumatic carditis in children. Am Heart J 1947;33:14-26.
Return to Text

It was our intention to present a reasonable and sound approach to the everyday situation faced by physicians reading routine electrocardiograms from newborns. In addition, we hoped to stimulate interest in the development of an inexpensive and easily applied screening tool that could be widely available to those caring for newborns and infants.

We agree with the main point of Yoldi et al. that "the great majority of neonates with a prolonged QTc will not be victims of SIDS." However, as we pointed out in our editorial, using a measurement of 440 msec as Schwartz et al. did in their study may lead to an inordinate number of false positives and the potential for overtreatment of a large number of infants. In the last paragraph of their article, the authors noted that "100 infants would have to be treated in order to save 2 lives" because of the low predictive value of a prolonged QT interval. It is our contention that to do so would be inappropriate and that the physician needs to decide which infants with a QTc of 440 msec or more should be treated and which are best served by a follow-up electrocardiogram or other test such as Holter monitoring.

What Schwartz and his coworkers have provided is a fairly strong argument to link prolongation of the QTc to SIDS. This is a problem that is devastating to the families in which SIDS has occurred and a major source of frustration and sorrow for the physicians who are involved with these families. With the additional help of molecular genetic techniques, we now appear to have a framework on the basis of which we can refine methods to identify the 2 of 100 infants who can be saved and thus affect a disease that has eluded our best efforts at prevention.

Richard A. Friedman, M.D.
Jeffrey A. Towbin, M.D.
Baylor College of Medicine
Houston, TX 77030