To the Editor: My coauthors and I wish to correct data reportedin our article on inhaled nitric oxide in preterm infants undergoingmechanical ventilation (July 27, 2006).1 During further datareview, we found that one infant in the placebo group in whombronchopulmonary dysplasia developed was incorrectly categorizedas surviving without chronic lung disease. Nine additional infantswere part of multiple births and thus received by design thesame treatment as the earlier enrolled sibling. Because siblingsshare both genetic and in utero environmental factors relatedto the occurrence of chronic lung disease, they represent clustereddata. Accordingly, we reanalyzed our data by means of the multipleoutputation approach used in the original analysis and obtaineda P value of 0.03 (relative benefit of nitric oxide in the overallpopulation, 1.26; 95% confidence interval [CI], 1.02 to 1.55)rather than a P value of 0.04 (relative benefit, 1.23; 95% CI,1.01 to 1.51), as reported in our article (in the Abstract,Table 2, and the third paragraph of the Results section). Analysisby means of the generalized-estimating-equations method,2 whichis also used for analysis of clustered data, gives a P valueof 0.03 (relative benefit, 1.45; 95% CI, 1.03 to 2.04). In Table2, the P value for survival among infants with a birth weightof 500 to 799 g should be 0.07 (relative benefit, 1.26; 95%CI, 0.98 to 1.62) rather than 0.14 (relative benefit, 1.20;95% CI, 0.94 to 1.54). In the first paragraph of the Resultssection, the last sentence should read, "The 582 remaining infantshad been delivered by 538 mothers (44 infants were assignedto a sibling's treatment)." In the third paragraph of the Resultssection, the first sentence should read, "In the group thatwas assigned to receive inhaled nitric oxide, 129 of 294 infantssurvived to 36 weeks of postmenstrual age without bronchopulmonarydysplasia (43.9 percent), as compared with 105 of 288 infantsin the placebo group (36.5 percent)." The second sentence ofthe sixth paragraph of the Results section (related to datareported in Table 5) should read, "We observed a significantinteraction between the age at study entry and treatment (P=0.005),"and the sixth sentence should read, "The effect of inhaled nitricoxide did not differ significantly according to race or ethnicgroup (P=0.06)" rather than "The effect of inhaled nitric oxideappeared to differ according to race or ethnic group (P=0.05)."These and other minor changes to the tables are listed in theSupplementary Appendix, available with the full text of thisletter at www.nejm.org.
Roberta A. Ballard, M.D. University of California at San Francisco San Francisco, CA 94143 ballardr{at}peds.ucsf.edu
Since publication of the article, Dr. Ballard reports havingreceived an honorarium and grant support from IKARIA/Inotherapeutics.No other potential conflict of interest relevant to this letterwas reported.
References
Ballard RA, Truog WE, Cnaan A, et al. Inhaled nitric oxide in preterm infants undergoing mechanical ventilation. N Engl J Med 2006;355:343-353. [Free Full Text]
Liang K-Y, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13-22. [Free Full Text]
Ballard, P. L., Truog, W. E., Merrill, J. D., Gow, A., Posencheg, M., Golombek, S. G., Parton, L. A., Luan, X., Cnaan, A., Ballard, R. A.
(2008). Plasma Biomarkers of Oxidative Stress: Relationship to Lung Disease and Inhaled Nitric Oxide Therapy in Premature Infants. Pediatrics
121: 555-561
[Abstract][Full Text]
Previous
