 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Peter G. Davis, MD
Colin J. Morley, MD
Division of Newborn Services
Royal Women's Hospital
Melbourne 3053, Australia
Andy McPhee, MD
Department of Neonatology
Women's and Children's Hospital
Adelaide 5006, Australia
John B. Carlin, PhD
Paediatrics,
University of Melbourne,
Melbourne 3052, Australia,
Clinical Epidemiology and Biostatistics Unit
Murdoch Childrens Research Institute
Melbourne 3052, Australia
Chief Investigators, DART study
We thank Dr Gordon for his comments about the DART (Dexamethasone: A Randomized Trial) study. He is concerned mostly about the problem of spontaneous intestinal perforation (SIP), which has been described in the randomized, controlled trials (RCTs) of postnatal corticosteroids, predominantly starting in the first week of life1 rather than later. The DART study differed in that although infants were eligible after the first week of life, the median age of starting was the fourth week of life, and we observed no SIP in either group. Because of the small sample size, we recognize that the confidence interval for the rate of zero is quite wide; the upper 95% confidence limit was 8.2% for the 35 infants in the dexamethasone group. Because of this, we have not said much about SIP, but we disagree with Dr Gordon that we "cannot say anything meaningful about SIP." We recorded the presence or absence of SIP, given that it was responsible for early termination of several RCTs.2,3
The focus of the DART study was to investigate the effects of postnatal steroids on survival and long-term neurosensory disability. These end points are more important than SIP alone, although we recognize that SIP could lead to either outcome in some infants. We suggest that any future RCT should focus on these end points and not on SIP as the major end point. We disagree that large-scale RCTs in neonatology are not possible; we have participated in the recently completed RCT of caffeine for apnea of prematurity (CAP). In CAP, >2000 infants were recruited in 5 years in many different centers around the world, and the major end point was similar to that of the DART study: survival free of major neurosensory disability. If survival free of major disability is significantly improved in any RCT, then the rate of SIP as part of such a study is not as important even if it is increased by the therapy under investigation.
We agree with Dr Gordon that co-interventions, such as use of indomethacin, can influence the results of any RCT. Dr Gordon suggests one strategy to deal with indomethacin in any future RCT: exclude it from the study. There are other, more practical strategies, however. One is to ensure that the treatment under investigation is blinded to the clinicians (as in the DART study), record the frequency of any co-interventions, and adjust for any imbalance in subsequent analysis of the data if, indeed, the co-interventions are important confounders.
The DART study did not provide the definitive answer to the problem of postnatal corticosteroid therapy; it was too small to do so. Because corticosteroids continue to be prescribed outside the context of clinical trials, we still believe that "our study helps reopen the debate on corticosteroids and supports future RCTs of low-dose dexamethasone designed to improve long-term rates of survival free of disability among infants beyond the first 1 week of life."4
REFERENCES
Related articles in Pediatrics:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
