Impact of a Physiologic Definition on Bronchopulmonary Dysplasia Rates

Study Population

From October 2000 to June 2002, 2582 consecutive inborn neonates (501–1249 g birth weight) were followed prospectively at the 17 centers of the National Institute of Child Health and Human Development Neonatal Research Network. The mean birth weight was 899 ± 212 g (mean ± SD) and postmenstrual age was 27 ± 2.4 weeks. At 36 ± 1 week's postmenstrual age, respiratory outcomes were defined in survivors as “BPD” or “no BPD” using 2 different definitions. The first definition of BPD used was a traditional clinical definition based on oxygen administration at 36 weeks' postmenstrual age. The second definition was a new physiologic definition based on combined measurement of oxygen saturation and oxygen administration. One hundred twenty-five infants with incomplete data were excluded from the analyses; 354 infants had died before 36 weeks' postmenstrual age, 151 infants were transferred (53 in oxygen), and 354 infants were discharged (12 in oxygen) before 36 weeks' postmenstrual age. This yielded a total cohort of 1598 in whom the physiologic definition was compared with the clinical definition.

Procedure for the Physiologic Definition of BPD

For this study, BPD was defined on a physiologic basis that combined oxygen and ventilation support with an assessment of saturation in selected infants that we previously reported.¹³ Briefly, at 35 to 37 weeks' postmenstrual age, infants treated with mechanical ventilation, continuous positive airway pressure, or supplemental oxygen concentration ≥30% and oxygen saturations between 90% and 96% were diagnosed with BPD without additional testing. Infants in supplemental oxygen <30% at rest with oxygen saturations between 90% and 96% or ≥30% with saturations >96% underwent a timed stepwise reduction to room air. For infants receiving oxygen by hood, oxygen was weaned in 2% increments. For infants receiving oxygen by nasal cannulae, flow was weaned initially in increments (for flow of 1.0–2.0: step down 0.5 liters per minute [lpm]; for flow 0.1-0.99 lpm: step down 0.1 lpm), and then oxygen concentration was reduced in increments of 20% to room air. Cannulae were removed from the nares for the remainder of the challenge. Oxygen that was given only during feedings was not included for the purposes of eligibility. Those who failed the reduction were diagnosed with BPD. No BPD was defined as treatment with room air with oxygen saturation ≥90% or passing a timed, continuously monitored oxygen-reduction test. No formal assessment of sleep state was made. Neonates were monitored continuously with a cardiorespiratory monitor and pulse oximeter (Nellcor, Mallinckrodt Inc, Hazelwood, MO) and observed directly by a trained neonatal research nurse throughout the reduction test. Values for heart rate, respiratory rate, oxygen saturation, frequency of apnea (cessation of breathing for >20 seconds), and bradycardia (heart rate <80 beats per minute for >10 seconds) were recorded every 60 seconds for a 15-minute baseline period. All occurrences of movement artifact (defined as visible motion of the infant and loss of the monitor's plethysmograph signal) were recorded, and corresponding saturation values were deleted. The test was performed in 4 parts: baseline, reduction phase, room-air observation phase, and return to usual oxygen. When piloting the physiologic definition, we arbitrarily chose the period of each oxygen-weaning step as 10 minutes and the period in room air as 60 minutes, allowing sufficient time at each step for the saturation to equilibrate. In the pilot data, saturation equilibrated by 5 minutes in the weaning step, and all infants who failed in the room-air phase did so before 30 minutes.¹³ Therefore, in the current study, each weaning step was shortened to 5 minutes and the period of observation in room air was decreased to 30 minutes.

From October 2000 to May 2001, failure was defined as saturation <88% for 5 consecutive minutes. Twenty-eight infants were studied using this definition. After this initial pilot phase, failure criteria were changed to improve the safety of infants studied by limiting exposure to extreme desaturation events. Beginning in June 2001, failure was defined as oxygen saturation 80% to 89% for 5 consecutive minutes or <80% for 15 seconds. There were 2 methods for passing during the room-air observation phase: a rapid pass and a pass after full monitoring. Rapid-pass criteria were met by successfully weaning to room air with all saturation ≥96% for 15 minutes. If the saturations were 90% to 95%, the infant was monitored for 30 minutes in room air and defined a pass when all saturations exceeded 90% in that 30-minute period.

Delivered oxygen concentration was measured directly in infants receiving oxygen by hood using an oxygen analyzer placed directly above the infant's head inside the hood. Among infants receiving oxygen by nasal cannula, the delivered oxygen concentration (termed the effective oxygen concentration) was calculated by using the technique described by Benaron and Benitz as modified for the Supplemental Therapeutic Oxygen for Prethreshold Retinopathy of Prematurity (STOP-ROP) trial.^10,14 This technique uses weight, oxygen liter flow, and oxygen concentration to calculate an effective concentration delivered by nasal cannula.

The attending physician individualized the neonate's management. Some infants received oxygen by hood and some by nasal cannula at the discretion of the clinical care team. Infants cared for in nasal cannula received oxygen delivered by either blended oxygen (15 centers) or 100% oxygen (2 centers) at variable flow rates (range: 0.01-2.00 lpm). In 6 centers, ≥1 infants were treated with nasal cannula with room air and variable flow (n = 16; range of flow: 0.03-2.00); all these infants were tested with a reduction test.

The clinical definition of BPD was assessed at 36 weeks' and 0 days' postmenstrual age and assigned in any neonate receiving oxygen supplementation for any portion of the day.

Safety Evaluation

Adverse events, defined before the study began, were collected in every infant and consisted of apnea (cessation of breathing >20 seconds), bradycardia (heart rate <80 beat per minute for >10 seconds), and increase in baseline fraction of inspired oxygen (Fio₂) of 5% persisting for >1 hour after the reduction test. Each infant was evaluated for adverse events in the 1 hour before the oxygen-reduction challenge and the 1 hour after the reduction. In addition, if any event occurred together with a saturation 80% to 89% for 5 consecutive minutes or <80% for 15 seconds during the challenge, the test was declared a failure and the infant was returned to the prior oxygen level immediately.

Human Subject Protections

The institutional review boards (IRBs) of all participating hospitals approved the study. Infants were studied after informed permission (written or verbal as required by the local IRB) was obtained from the parent or guardian. The risk of study participation was minimized by the continuous presence of an experienced neonatal research nurse. All infants were returned to their baseline supplemental oxygen at the end of the oxygen-reduction challenge. Results of the room-air challenge were handled differently at different centers in conjunction with local IRB requirements. In 12 centers, the results of the test were given to the primary physician caring for the infant; at the remaining 5 centers, results were concealed. The approach at an individual center was consistent throughout the study period. Additional modifications in treatment after the oxygen-reduction phase, if any, were made at the primary physician's discretion.

Statistical Analyses

Continuous variables were described by mean and SD. Group differences on continuous measures were compared with Student's t tests/analysis of variance. Categorical variables were assessed with frequency distribution, and group differences were compared with the χ² test. The McNemar test was used for comparing BPD rates by clinical definition and physiologic definition and for differences in the rates of medication use. The impact of baseline effective Fio₂ and saturation on oxygen-reduction test results was assessed with regression models to adjust for potential confounding variables. A posthoc power analysis was done based on the McNemar statistic.¹⁵ The sample size was sufficient to detect an absolute rate difference of 6% at 80% power and type I error of 0.05.

Respiratory Outcomes

The outcomes of the neonates at 36 weeks' postmenstrual age are shown in Fig 1. The physiologic definition diagnosed BPD in 398 (25.0%) of 1598 neonates. Of these 398 patients, 272 were diagnosed with BPD because of their continued use of ventilator, continuous positive airway pressure, or oxygen support ≥30% and 126 because of a failed oxygen-reduction test. In contrast, the clinical definition diagnosed BPD in 560 (35.0%) of these 1598 infants. Overall, using the physiologic definition reduced the number of infants diagnosed with BPD from 35% to 25% (P < .0001). The physiologic definition also modestly reduced the variation among centers (variation: 15–66% with the clinical definition and 9–57% with the physiologic definition; P < .0001), although substantial variation remained.

• Download figure
• Open in new tab
• Download powerpoint

Fig 1

Respiratory outcomes of the physiologic definition of BPD at 36 weeks' postmenstrual age. CPAP indicates continuous positive airway pressure.

During the physiologic-definition stage, 227 infants were studied with an oxygen-reduction test at a mean of 36.4 ± 3.4 weeks' postmenstrual age. At the time of the oxygen-reduction test, 11 of the infants were receiving oxygen by hood, and 216 (92%) were receiving oxygen by nasal cannula with flow rates at a mean of 0.49 lpm (range: 0.02–2.00 lpm). Of 227 infants tested, 101 (44%) passed the oxygen-reduction phase and were successfully weaned to room air. Of the 126 infants who failed, 83 (66%) failed during the reduction phase, 39 failed during the room-air phase, and 4 had incomplete information on the time of failure. The mean duration of the oxygen-reduction test was 34 ± 24 minutes. As expected, the duration of the study was significantly shorter in those who failed compared with those who passed (23 ± 16 vs 49 ± 24 minutes; P < .001). Infants who failed during the room-air phase did so at a mean of 20 ± 11 minutes.

Both the incidence of BPD and the magnitude of the impact of the physiologic definition on BPD rates varied by center (P < .0001; Fig 2). The overall reduction was 10%, with a range of 0% to 44%. Of 17 centers, 16 had a decrease in their BPD rate and 1 remained the same. The center with the largest impact had a reduction in the BPD rate from 60% with the clinical definition to 16% with the physiologic definition. This center routinely supplements oxygen to achieve saturations >98% in patients with stage 2 or greater retinopathy of prematurity.

• Download figure
• Open in new tab
• Download powerpoint

Fig 2

Comparison of BPD rate by the clinical definition and physiologic definition of BPD at each center. The overall reduction in mean BPD rate (P < .0001) and the reduction in variation (P < .01) between centers are both significant.

For the infants who had the oxygen-reduction test performed, we compared those who passed and those who failed to determine if test failure could be predicted by a preexisting demographic characteristic (Table 1). Infants who failed were of similar birth weight and gestation to those who passed. Baseline heart rate and respiratory rate were also similar among those who failed and those who passed. Saturation before the reduction was significantly higher in those who passed compared with those who failed (97.0 ± 2.3 vs 95.8 ± 2.8; P < .001), and the amount of oxygen supplementation was significantly less (23% ± 3% vs 26 ± 5%; P < .001).

Safety of the Oxygen-Reduction Test

To assess the safety of the oxygen-reduction challenge, we assessed the frequency of predefined potential adverse events in every infant in the 1 hour before and 1 hour after the challenge. Results of the safety monitoring are shown in Table 2. The number of infants with apnea or bradycardia was not different in the periods before and after the challenge. Desaturation events, defined as saturation <88% for >5 consecutive minutes, did occur in more infants after the challenge (3 vs 15 infants; P < .0012). However, the frequency of increased oxygen support in the period before and after the challenge was not different, suggesting that the events were not viewed as significant by the clinical team caring for the infant. No infant had ventilation instituted after the challenge.

Validity of the Physiologic Definition as a Measure of Pulmonary Morbidity

To further assess the validity of the physiologic definition, we correlated the definition with other markers of ongoing pulmonary disease and health-resource use. Infants diagnosed with BPD by the physiologic definition compared with those with no BPD were more likely to remain in hospital at 120 days of age (44.7% vs 6.6%; P < .0001), and more likely to be discharged from the hospital in oxygen (53.3% vs 3.7%; P < .0001). Medication use at 36 weeks' postmenstrual age was also higher in those diagnosed with BPD (Table 3). The frequency of medication use did vary by center (P < .001).

We further analyzed the subset of 227 who received the oxygen-reduction challenge. As previously shown, infants who passed and infants who failed the oxygen-reduction challenge were of similar birth weight, postmenstrual age, and age at the time of study (Table 1). Diuretic use at 36 weeks (39% vs 49%; P = .11) and steroid use at 36 weeks (2% vs 3%; P = .58) were not different between those who passed and those who failed. However, significantly more of those who failed the room-air challenge were discharged from the hospital in oxygen than in those who passed (58% vs 26%; P < .001) The physiologic definition correlates with ongoing pulmonary morbidity and health-resource utilization.

Potential Impact of Varying Saturation Cutoff for Failure

Because the optimal oxygen-saturation value for convalescent preterm infants is not known, we assessed the impact of varying oxygen-saturation cutoff values between 88% and 92% on the test results. The failure criteria of oxygen saturation <90% for 5 consecutive minutes or saturation <80% for 15 seconds proved to be a robust cutoff point. When we assessed the potential impact of lowering the failure criteria to <89% for 5 consecutive minutes (while still failing any <80% for 15 seconds), only 1 additional infant passed the challenge. No additional infants passed at a failure cutoff of <88% for 5 consecutive minutes. Similarly, when we assessed higher cutoff values of 91% and 92% for a failure cutoff, no additional infants failed the test.

Advisory Committee

M. C. Walsh, MD, MS, and A. A. Fanaroff, MD (Case Western Reserve University, Cleveland, OH); A. H. Jobe, MD, PhD (University of Cincinnati, Cincinnati, OH); R. Higgins, MD (National Institute of Child Health and Human Development, Bethesda, MD); N. Finer, MD (University of California, San Diego, CA); and K. Poole, PhD (Research Triangle Institute, Research Triangle Park, NC).

Intervention Centers

N. Kazzi, MD, K. Hayes-Hart, RN, M. Betts, RRT, S. Shankaran, MD, principal investigator, and G. Muran, RN (Wayne State University, Detroit, MI); A. Laptook, MD, M. Martin, RN, and J. Allen, RRT (University of Texas Southwestern, Dallas, TX); W. A. Engle, MD, L. Miller, RN, R. Hooper, RRT, and J. Lemons, MD, principal investigator (Indiana University, Indianapolis, IN); W. Rhine, MD, C. Kibler, RN, J. Parker, RRT, D. Stevenson, MD, principal investigator, and B. Ball, BS (Stanford University, Palo Alto, CA); M. Rasmussen, MD, M. Grabarczyk, BSN, C. Joseph, RRT, and K. Arnell, BSN (Sharp Mary Birch Hospital for Women, San Diego, CA); G. Heldt, MD, R. Bridge, RN, J. Goodmar, RRT, N. Finer, MD, and C. Henderson, RCP (University of California, San Diego, CA); and S. Buchter, MD, M. Berry, RN, I. Seabrook, RRT, B. Stoll, MD, principal investigator, and E. Hale, RN (Emory University, Atlanta, GA).

Benchmark Centers

S. Duara, MD, and R. Everette, RN (University of Miami, Miami, FL); W. Carlo, MD, and M. Collins, RN (University of Alabama, Birmingham, AL); and W. Oh, MD, and A. Hensman, RN (Brown University, Providence, RI).

Control Centers

M. T. O'Shea, MD, MPH, and N. Peters, RN (Wake Forest University, Winston-Salem, NC); J. Tyson, MD, MPH, and G. McDavid, RN (University of Texas, Houston, TX); A. A. Fanaroff, MD, M. C. Walsh, MD, MS, and N. Newman, RN (Case Western Reserve University, Cleveland, OH); D. Phelps, MD, and L. Reubens, RN (University of Rochester, Rochester, NY); R. A. Ehrenkranz, MD, and P. Gettner, RN (Yale University, New Haven, CT); C. Michael Cotten, MD, and K. Auten, RN (Duke University, Durham, NC); and E. Donovan, MD, and C. Grisby, RN (University of Cincinnati, Cincinnati, OH).

Statistical Center

Qing Yao, PhD, and Ken Poole, PhD (Research Triangle Institute, Research Triangle Park, NC).