eLetters to:

ARTICLE: Linda J. Van Marter, Elizabeth N. Allred, Marcello Pagano, Ulana Sanocka, Richard Parad, Marianne Moore, Mervyn Susser, Nigel Paneth, Alan Leviton, the Neonatology Committee, and for the Developmental Epidemiology Network Do Clinical Markers of Barotrauma and Oxygen Toxicity Explain Interhospital Variation in Rates of Chronic Lung Disease? Pediatrics 2000; 105: 1194-1201 [Abstract] [Full text] [PDF]

eLetters: Submit a response to this article

eLetters published:

Untitled

Keith J Barrington   (5 July 2000)

Untitled 5 July 2000
Keith J Barrington, Neonatologist McGIll University Send letter to journal: Re: this article kbarri{at}po-box.mcgill.ca Keith J Barrington	The provocative paper by Van Marter and colleagues presents extensive and, might I say, "fancy" statistics to determine correlates of the differing incidence of bronchopulmonary dysplasia between 2 centers. All of these analyses of course are predicated on the assumption that there is a real difference in lung injury between the two centers, and that this relates to management practices, as well as to patient differences. Unfortunately the use of oxygen at 36 weeks post conceptional age is by no means an objective measure of lung injury. Such oxygen use is critically dependent on management "style". One center might easily continue oxygen therapy for an infant who exhibits a saturation of 92% during feeds, whereas another would discontinue therapy unless the infant desaturates to below 90% at rest, for example. The first infant would be classed as having BPD whereas the second (who may have more severe lung injury) would be classed as not having BPD. No statistical analysis could correct for such a difference, yet no mention of the difficulties in truly assessing the occurrence of significant lung injury is mentioned in the paper. Much of the literature on BPD has been muddied by treating the condition as if it was an either/or situation. An infant who has oxygen therapy discontinued at 35 weeks and 6 days is not necessarily significantly worse off than an infant who receives oxygen for 2 further days! Although this is self-evident it is ignored by analyses such as those contained in this paper. Van Marter et al have demonstrated a differing incidence of the use of oxygen at 36 weeks postconceptional age between centers, the implications of that can only be determined once it is certain that there is truly an important difference in the lung injury between the centers, and that such a difference if it exists is truly associated with differing management practices.