- ECG =
- electrocardiogram •
- SIDS =
- sudden infant death syndrome
The article by Schwartz et al1 has major implications if the recommendations of the authors are put into place by the medical establishment. In their conclusions the authors recommend that “neonatal electrocardiographic screening may permit the early identification of a substantial percentage of infants at risk with SIDS and the institution of preventative measures may therefore be possible.”
Given the implications of these conclusions, it is important to examine with extreme care the data in this article. The most important concerns relate to methodology. Because electrocardiograms (ECGs) were recorded on 34 442 newborn infants born in 9 maternity hospitals, I would like to have seen a lot more information about the epidemiologic nature of this database. For example, what percentage of ascertainment was obtained and why, if any, were infants excluded from each of the 9 centers involved?
Turning to the question of the ECG itself, I am worried about the lack of information with respect to paper speed quality and the reproducibility and repeatability of measurements. There is no statement in this article that measurements were made in a blinded fashion. The measurement of QT intervals on the ECGs of newborn infants is subject to considerable variability. I note also, that the longest QTc measurement identified in a sudden infant death syndrome (SIDS) victim was identical in value (QTc = 563; patient 4) to that reported in an article published in 1982.2 This suggests that the authors have not taken all the ECGs on infants who have suffered SIDS from their total sample and subjected them with controls to a blinded analysis. In my view, if such an analysis has not yet been undertaken, it now should be and the results published.
I do agree with the authors that it was a good idea to divide the RR intervals into 17 categories.
I am also concerned about the nature of the postmortem examinations undertaken on the patients in this study. In an editorial by Guntheroth in 19823 there was a note that some of the infants in the earlier studies of this group had not undergone a postmortem investigation. The diagnosis of SIDS cannot be made in the absence of an autopsy. Neurologic abnormalities as well as biochemical disorders in the newborn period can prolong QT intervals and therefore neurologic and other abnormalities must be excluded by an adequate postmortem before a diagnosis of SIDS can be made. The authors should provide us with more information about the postmortems performed.
Finally, Schwartz et al then point out that 6 of the infants in our own prospective study had QTc values exceeding the 90th centile for the study population.4 This is not true; only 4 of the 15 SIDS patients had QTc values that exceeded the 90th centile and 2 were on it.
- Received October 29, 1998.
- Accepted October 29, 1998.
Reprint requests to (D.P.S.) North Staffordshire Hospital, Department of Academic Paediatrics, Stoke-on-Trent, Staffordshire, England ST4 6QG.
- Schwartz PJ,
- Montemerlo M,
- Facchini M,
- et al.
- Guntheroth WG
- Southall DP,
- Arrowsmith WA,
- Stebbens V,
- Alexander JR
- Copyright © 1999 American Academy of Pediatrics