PEDIATRICS Vol. 108 No. 2 August 2001, pp. 525

Barotrauma, Oxygen Toxicity, and Chronic Lung Disease

To the Editor.

We read the article by Van Marter et al¹ with great interest. The authors present results of an epidemiologic study comparing the pulmonary outcomes of infants born and managed in New York and in Boston. The main outcome indicator was the requirement for supplementary oxygen at 36 weeks' postmenstrual age. They found marked differences in this outcome, with fewer infants requiring oxygen at this time in New York. A number of interesting correlates are discussed, including the observed differences in intubation rates, surfactant use, and mechanical support other than continuous positive airway pressure. All in all, the study lends more credence to a respiratory treatment style of intervention only when forced, allowing the infants maximum opportunity to perform whatever breathing work they cana style that might be called the "leave well enough alone" technique.

A major problem with this study is the use of the "oxygen at 36 weeks" surrogate for chronic lung disease (CLD). This article, like others,^2-4 is flawed in its failure to explain how the various treating physicians assessed the infants' ongoing need for supplemental oxygen and, most importantly, how they decided to withdraw supplemental oxygen therapy. In a single institution, one can assume that infants in oxygen at 36 weeks have worse lung disease than those in room air. It does not follow that infants in oxygen at one institution are similar to those at another unless the oxygen weaning strategies are the same. In addition, the results of the recent "STOP-ROP" trial,⁵ one of the few studies to clearly define the oxygen use criteria during the study period, demonstrates that infants maintained at higher oxygen saturation values need supplemental oxygen for longer periods of time and may have worse pulmonary outcomes. The decision to discontinue supplemental oxygen, according to our very informal survey of colleagues, is based on a variety of factors and is made differently by different physicians. Did these authors have an agreed-upon oxygen weaning strategy? If so, the article's conclusions are greatly strengthened. If not, I suggest we find a way to focus on this issue if we wish to continue using this indicator as a measure of the CLD rate in our populations.

Since publication of the article by Avery et al⁶ describing very different rates of CLD in 8 prominent institutions, those of us who care for newborns have struggled to understand the complex development of CLD in the premature newborn. How we use the mechanical ventilators available to us clearly plays a role in this process. How we use oxygen, both to treat and to define this problem, may, in fact, obscure it.

Mark C. Mammel, MD
Raye-Ann O. deRegnier, MD
Children's Hospital-St Paul
St Paul, MN 55102

REFERENCES

1. Van Marter LJ, Allred EN, Pagano M, Do clinical markers of barotrauma and oxygen toxicity explain interhospital variation in rates of chronic lung disease? Pediatrics. 2000; 105:1194-1201 [Abstract/Free Full Text]
2. Fanaroff AA, Wright LL, Stevenson DK, Very-low-birth-weight outcomes of the National Institute of Health and Human Development Neonatal Research Network, May 1991 through December 1992. Am J Obstet Gynecol. 1995; 173:1423-1431 [CrossRef][Medline]
3. Knight DB, Liggins GC, Wealthall SR A randomized, controlled trial of antepartum thyrotropin-releasing hormone and betamethasone in the prevention of respiratory disease in preterm infants. Am J Obstet Gynecol. 1994; 171:11-16 [Medline]
4. Tapia JL, Ramirez R, Cifuentes J, The effect of early dexamethasone administration on bronchopulmonary dysplasia in preterm infants with respiratory distress syndrome. J Pediatr. 1998; 132:48-52 [CrossRef][Medline]
5. The STOP-ROP, Multicenter Study Group Supplemental therapeutic oxygen for prethreshold retinopathy of prematurity (STOP-ROP), a randomized, controlled trial. I: primary outcomes. Pediatrics. 2000; 105:295-310 [Abstract/Free Full Text]
6. Avery ME, Tooley WH, Keller JB, Is chronic lung disease in low birth weight infants preventable? A survey of eight centers. Pediatrics. 1987; 79:26-30 [Abstract/Free Full Text]

To the Editor.

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 2 centers, and that this relates to management practices, as well as to patient differences. Unfortunately the use of oxygen at 36 weeks' postconceptional 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 classified as having bronchopulmonary dysplasia (BPD) whereas the second (who may have more severe lung injury) would be classified 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 additional days. Although this is self-evident, it is ignored by analyses such as those contained in this article.

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.

Keith J. Barrington, MB, ChB
Department of Pediatrics
McGill University
Montreal, Quebec H3A 1A1 Canada

In Reply.

Drs Barrington, Mammel, and deRegnier make excellent points in their critique of our article. We agree with their main criticism: selecting as an outcome measure for our study an objective pathologic or physiologic outcome measure would have been preferable to a clinical endpoint, such as the requirement for supplemental oxygen at a specific age. Dr Barrington correctly added that even a continuous measure of oxygen administration is a more robust measure than is a single measure of oxygen exposure.

We chose to use treatment with supplemental oxygen at 36 weeks' postmenstrual age because it was consistent with the most current clinical definition of chronic lung disease (CLD) among surviving preterm infants.¹ Unlike the previous definition of oxygen therapy at 28 days' postnatal age, the new definition incorporates an adjustment for the infant's degree of immaturity at birth. As previously noted, for CLD radiographic criteria are superior to clinical ones.^2,3 CLD definitions based on requirement for supplemental oxygen at a specific postnatal or gestational age are flawed because they are susceptible to influence by practices and policies that vary among clinicians and institutions. At the outset of our study, we believed the thresholds for oxygen administration and weaning were consistent between the institutions. However, variation in monitoring also could affect differential rates of supplemental oxygen use. Having failed to collect data documenting consistency in monitoring and intervention thresholds, a possible alternative explanation for our study results is that our findings simply reflect variability in 1 or both of these factors. On the other hand, it seems unlikely that variability among the units in 1 or both of these factors would explain the fourfold increase in CLD risk we observed. Furthermore, we thoroughly evaluated our previous hypotheses from a number of perspectivesincluding evaluating CLD risk as a function of duration of mechanical ventilationand found the results to be similar. There is no question that, comparing the 2 centers, there was a consistent pattern in favor of lower use of and more rapid weaning from mechanical ventilation at Babies' Hospital.

A large, well-designed clinical trial might be called for to answer the question of whether the hypothesis raised by our observational study is valid and rates of CLD among preterm infants could be reduced by gentler and less invasive methods of ventilatory support. In the interim, the points raised by these 2 critiques are well-taken and our study results should be interpreted with caution.

Linda J. Van Marter, MD, MPH^{*, , **}
Elizabeth N. Allred, MS^*,
Marcello Pangano, PhD^*,
Ulana Sanocka, MD^{, ¶}
Richard Parad, MD, MPH^{*, , **}
Marianne Moore, BA^{*, , **}
Mervyn Susser, MB, BCh
Nigel Paneth, MD^§
Alan Leviton, MD, MS^*,
* Children's Hospital
Boston, MA
 Harvard Medical School
Boston, MA
^§ Michigan State University
East Lansing, MI
 Columbia University
New York, NY
^¶ Babies' and Children's Hospital
New York, NY
^# St Luke's-Roosevelt Medical Center
New York, NY
^** Brigham and Women's Hospital
Boston, MA
 Harvard School of Public Health
Boston, MA

REFERENCES

1. Vohr BR, Wright LL, Dusick AM, Neurodevelopmental and functional outcomes of extremely low birth weight infants in the National Institute of Child Health and Human Development Neonatal Research Network, 1993-1994. Pediatrics. 2000; 105:1216-1226 [Abstract/Free Full Text]
2. Merritt TA, Northway W Jr, Boynton BR, Edwards DK, Hallman M, Berry C The BPD problem. Pediatrics. 1991; 88:189-191 [Abstract/Free Full Text]
3. Palta M, Sadek M, Barnet JH, Evaluation of criteria for chronic lung disease in surviving very low birth weight infants. Newborn Lung Project. J Pediatr. 1998; 132:57-63 [CrossRef][Medline]