Journal of Perinatology (2014) 34, 59–63 & 2014 Nature America, Inc. All rights reserved 0743-8346/14 www.nature.com/jp
ORIGINAL ARTICLE
Infants born at o29 weeks: pulmonary outcomes from a hybrid perinatal system WE Truog, MF Nyp, J Taylor, LL Gratny, H Escobar, WM Manimtim, CI Lachica, A Khmour, OO Oluola, AA Oshodi, M Norberg, H Dai and EK Pallotto OBJECTIVE: To assess pulmonary outcomes of infants o29 weeks gestational age (GA), delivered at level I, II and III facilities, to identify potentially modifiable factors affecting bronchopulmonary dysplasia (BPD) severity and to assess the external generalizability of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) BPD Outcome Estimator. STUDY DESIGN: Outcomes for infants o29 weeks GA born during (2008–2010) and delivered either at an inborn level III center or in a level II or III metropolitan area hospital with transfer to a level IV center, or delivered in a distant level I or II center and then transported to a level IV center were assessed. BPD severity was compared with the NICHD Neonatal BPD Outcome Estimator. RESULT: Of 158 infants who comprised the cohort, 28 (17.8%) had no BPD, 39 (24.2%) had mild BPD, 45 (28.7%) had moderate BPD, 31 (19.7%) had severe BPD and 15 (9.6%) died at p36 weeks post menstrual age. Site of birth did not predict severe BPD or death. Receiver operator characteristic curves showed fair predictability for none/mild and severe BPD. CONCLUSION: BPD severity was not dependent on site of birth. The NICHD BPD outcome estimator provides fair prediction for extreme outcomes. Journal of Perinatology (2014) 34, 59–63; doi:10.1038/jp.2013.125; published online 17 October 2013 Keywords: bronchopulmonary dysplasia; prematurity; pulmonary outcomes
INTRODUCTION Hospitals with variable resources for perinatal resuscitation and stabilization of very preterm infants continue to deliver extremely preterm infants o29 weeks gestation. In all, 74.7% of infants o1500 g birthweight were delivered in hospitals with high-level intensive care units.1 The goal of regionalizing perinatal care is to facilitate delivery of high-risk infants in high-level perinatal centers. However, the emergent or urgent nature of many of these deliveries renders this goal difficult to achieve. There have been no recent meaningful improvements in this goal.1 Outcome reports that utilize databases comprised only with infants inborn at level III or IV centers may exclude infants whose pulmonary outcomes could differ because of circumstances surrounding delivery. Bronchopulmonary dysplasia (BPD) remains a common outcome for infants o29 weeks gestation.2 As overall survival rates have either stabilized or increased since the mid-1990s,2,3 the incidence of BPD has not decreased.4,5 Recently tested interventions, for example, early continuous positive airway pressure,6 oscillating ventilators,7 noninvasive ventilation,8 have not been demonstrated to reduce the incidence or severity of BPD as measured physiologically. Laughon et al9 created a BPD outcomes estimator for extremely preterm infants by utilizing data obtained from the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network Benchmarking Trial.10 This trial evaluated outcomes at 17 centers. Infants between 23 and 30 weeks gestational age (GA) with a birthweight of 401–1250 g, who survived more than 12 h were included. Infants o23 weeks were excluded because they all developed BPD and infants 430 weeks were excluded because
less than 1% developed BPD. Others have demonstrated that outcomes at 36 weeks gestation correlate with very-long-term pulmonary outcomes, but may underestimate the impact of even mild forms of BPD for later pulmonary health.11,12 Given this background, we undertook this evaluation with three objectives: (1) to assess pulmonary outcomes at 36 weeks post menstrual age (PMA) of infants born at o29 weeks GA and delivered at facilities with three different levels of care within our Center’s expanded catchment area, (2) to identify potentially modifiable factors that could affect pulmonary outcome and (3) to assess the generalizability of the Eunice Kennedy Shriver National Institute Child Health and Human Development (NICHD) BPD Outcomes Estimator and create receiver operator characteristic (ROC) curves for predicting outcomes from the 7- and/or 14-day estimates. MATERIALS AND METHODS Data repository A data repository in the Center for Infant Pulmonary Disorders located at Children’s Mercy Hospitals and Clinics provided de-identified data for this study. The process, approved by the Pediatric Institutional Review Board at Children’s Mercy Hospitals and Clinics, allows comprehensive clinical data collection and storage of the medical results for all relevant infants and allows investigators to access data while maintaining de-identification.
Infant inclusion criteria There were 184 infants born alive o29 weeks during the 3-year period (2008–2010) at one of three types of locations. Group A includes infants
Department of Pediatrics, Children’s Mercy Hospitals and Clinics, Center for Infant Pulmonary Disorders, University of Missouri-Kansas City School of Medicine, Kansas, MO, USA. Correspondence: Dr WE Truog, Division of Neonatology, Department of Pediatrics, Children’s Mercy Hospitals and Clinics, Center for Infant Pulmonary Disorders, 2401 Gillham Road, Kansas City, MO 64108, USA. E-mail: [email protected] Received 3 June 2013; revised 15 August 2013; accepted 27 August 2013; published online 17 October 2013
Infants born at o29 weeks WE Truog et al
60 delivered at the inborn level III center (Truman Medical Center-Hospital Hill) with all deliveries staffed by neonatal faculty, fellows and neonatal nurse practitioners. Group B includes infants delivered in local metropolitan area hospitals (level II or III) with a neonatal faculty, neonatal nurse practitioner and neonatal newborn transport team present at delivery with immediate transfer to the level IV center (Children’s Mercy Hospital). Group B metropolitan area hospitals are o15 km or 30 min drive from the level IV NICU. Group C includes infants delivered in level I or II centers neonatal resuscitation program certified, with local pediatricians/family practitioners and when possible neonatal transport team present at the delivery and then immediately transferred to the level IV NICU by the dedicated neonatal transport team. Group C consisted of between 20 and 160 km from the level IV NICU. Infants with major malformations or those transferred from any site at X24 h of age or those dying within 12 h of life were excluded from further analysis (Figure 1).
BPD outcome We defined five outcome categories based on pulmonary status at 36±1 week PMA according to criteria established by Ehrenkranz et al13 and Walsh et al.5 None (no BPD): infants who required supplemental oxygen for less than 28 days to maintain their SpO2490% and then ambient air at 36 weeks or upon discharge from the NICU (whichever came first). Mild (BPD): infants who required supplemental oxygen for at least 28 days to maintain their SpO2490% and then ambient air at 36 weeks or upon discharge from the NICU. Moderate (BPD): infants who required supplemental oxygen less than 30% to maintain their SpO2490% at 36 weeks or upon discharge. Severe (BPD): infants who required supplemental oxygen greater than 30% or positive pressure ventilation to maintain their SpO2490% at 36 weeks or upon discharge. Death: infants who died 12 h after birth and before 36 weeks PMA. The effective O2 delivery (FEO2), calculated from available data, determined respiratory status at 36 weeks (Benaron and Benitz).14
Use of BPD estimator Data from the repository included: gestational age, birthweight, sex, race/ ethnicity, ventilator type (HF Vent, IMV/SIMV, continuous positive airway pressure, Cannula/Hood or None) and delivered FiO2 at day of life 7 and 14. The online Neonatal BPD Outcome Estimator (NICHD Neonatal Research Network at https://neonatal.rti.org/index.cfm?fuseaction=BPDcalculator.start was then used to calculate the percent likelihood of developing each of five outcomes at 36 weeks PMA (none, mild, moderate, severe, death). The infants with X60% likelihood of having none or mild BPD as the outcome
ORIGINAL ARTICLE
Infants born at o29 weeks: pulmonary outcomes from a hybrid perinatal system WE Truog, MF Nyp, J Taylor, LL Gratny, H Escobar, WM Manimtim, CI Lachica, A Khmour, OO Oluola, AA Oshodi, M Norberg, H Dai and EK Pallotto OBJECTIVE: To assess pulmonary outcomes of infants o29 weeks gestational age (GA), delivered at level I, II and III facilities, to identify potentially modifiable factors affecting bronchopulmonary dysplasia (BPD) severity and to assess the external generalizability of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) BPD Outcome Estimator. STUDY DESIGN: Outcomes for infants o29 weeks GA born during (2008–2010) and delivered either at an inborn level III center or in a level II or III metropolitan area hospital with transfer to a level IV center, or delivered in a distant level I or II center and then transported to a level IV center were assessed. BPD severity was compared with the NICHD Neonatal BPD Outcome Estimator. RESULT: Of 158 infants who comprised the cohort, 28 (17.8%) had no BPD, 39 (24.2%) had mild BPD, 45 (28.7%) had moderate BPD, 31 (19.7%) had severe BPD and 15 (9.6%) died at p36 weeks post menstrual age. Site of birth did not predict severe BPD or death. Receiver operator characteristic curves showed fair predictability for none/mild and severe BPD. CONCLUSION: BPD severity was not dependent on site of birth. The NICHD BPD outcome estimator provides fair prediction for extreme outcomes. Journal of Perinatology (2014) 34, 59–63; doi:10.1038/jp.2013.125; published online 17 October 2013 Keywords: bronchopulmonary dysplasia; prematurity; pulmonary outcomes
INTRODUCTION Hospitals with variable resources for perinatal resuscitation and stabilization of very preterm infants continue to deliver extremely preterm infants o29 weeks gestation. In all, 74.7% of infants o1500 g birthweight were delivered in hospitals with high-level intensive care units.1 The goal of regionalizing perinatal care is to facilitate delivery of high-risk infants in high-level perinatal centers. However, the emergent or urgent nature of many of these deliveries renders this goal difficult to achieve. There have been no recent meaningful improvements in this goal.1 Outcome reports that utilize databases comprised only with infants inborn at level III or IV centers may exclude infants whose pulmonary outcomes could differ because of circumstances surrounding delivery. Bronchopulmonary dysplasia (BPD) remains a common outcome for infants o29 weeks gestation.2 As overall survival rates have either stabilized or increased since the mid-1990s,2,3 the incidence of BPD has not decreased.4,5 Recently tested interventions, for example, early continuous positive airway pressure,6 oscillating ventilators,7 noninvasive ventilation,8 have not been demonstrated to reduce the incidence or severity of BPD as measured physiologically. Laughon et al9 created a BPD outcomes estimator for extremely preterm infants by utilizing data obtained from the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network Benchmarking Trial.10 This trial evaluated outcomes at 17 centers. Infants between 23 and 30 weeks gestational age (GA) with a birthweight of 401–1250 g, who survived more than 12 h were included. Infants o23 weeks were excluded because they all developed BPD and infants 430 weeks were excluded because
less than 1% developed BPD. Others have demonstrated that outcomes at 36 weeks gestation correlate with very-long-term pulmonary outcomes, but may underestimate the impact of even mild forms of BPD for later pulmonary health.11,12 Given this background, we undertook this evaluation with three objectives: (1) to assess pulmonary outcomes at 36 weeks post menstrual age (PMA) of infants born at o29 weeks GA and delivered at facilities with three different levels of care within our Center’s expanded catchment area, (2) to identify potentially modifiable factors that could affect pulmonary outcome and (3) to assess the generalizability of the Eunice Kennedy Shriver National Institute Child Health and Human Development (NICHD) BPD Outcomes Estimator and create receiver operator characteristic (ROC) curves for predicting outcomes from the 7- and/or 14-day estimates. MATERIALS AND METHODS Data repository A data repository in the Center for Infant Pulmonary Disorders located at Children’s Mercy Hospitals and Clinics provided de-identified data for this study. The process, approved by the Pediatric Institutional Review Board at Children’s Mercy Hospitals and Clinics, allows comprehensive clinical data collection and storage of the medical results for all relevant infants and allows investigators to access data while maintaining de-identification.
Infant inclusion criteria There were 184 infants born alive o29 weeks during the 3-year period (2008–2010) at one of three types of locations. Group A includes infants
Department of Pediatrics, Children’s Mercy Hospitals and Clinics, Center for Infant Pulmonary Disorders, University of Missouri-Kansas City School of Medicine, Kansas, MO, USA. Correspondence: Dr WE Truog, Division of Neonatology, Department of Pediatrics, Children’s Mercy Hospitals and Clinics, Center for Infant Pulmonary Disorders, 2401 Gillham Road, Kansas City, MO 64108, USA. E-mail: [email protected] Received 3 June 2013; revised 15 August 2013; accepted 27 August 2013; published online 17 October 2013
Infants born at o29 weeks WE Truog et al
60 delivered at the inborn level III center (Truman Medical Center-Hospital Hill) with all deliveries staffed by neonatal faculty, fellows and neonatal nurse practitioners. Group B includes infants delivered in local metropolitan area hospitals (level II or III) with a neonatal faculty, neonatal nurse practitioner and neonatal newborn transport team present at delivery with immediate transfer to the level IV center (Children’s Mercy Hospital). Group B metropolitan area hospitals are o15 km or 30 min drive from the level IV NICU. Group C includes infants delivered in level I or II centers neonatal resuscitation program certified, with local pediatricians/family practitioners and when possible neonatal transport team present at the delivery and then immediately transferred to the level IV NICU by the dedicated neonatal transport team. Group C consisted of between 20 and 160 km from the level IV NICU. Infants with major malformations or those transferred from any site at X24 h of age or those dying within 12 h of life were excluded from further analysis (Figure 1).
BPD outcome We defined five outcome categories based on pulmonary status at 36±1 week PMA according to criteria established by Ehrenkranz et al13 and Walsh et al.5 None (no BPD): infants who required supplemental oxygen for less than 28 days to maintain their SpO2490% and then ambient air at 36 weeks or upon discharge from the NICU (whichever came first). Mild (BPD): infants who required supplemental oxygen for at least 28 days to maintain their SpO2490% and then ambient air at 36 weeks or upon discharge from the NICU. Moderate (BPD): infants who required supplemental oxygen less than 30% to maintain their SpO2490% at 36 weeks or upon discharge. Severe (BPD): infants who required supplemental oxygen greater than 30% or positive pressure ventilation to maintain their SpO2490% at 36 weeks or upon discharge. Death: infants who died 12 h after birth and before 36 weeks PMA. The effective O2 delivery (FEO2), calculated from available data, determined respiratory status at 36 weeks (Benaron and Benitz).14
Use of BPD estimator Data from the repository included: gestational age, birthweight, sex, race/ ethnicity, ventilator type (HF Vent, IMV/SIMV, continuous positive airway pressure, Cannula/Hood or None) and delivered FiO2 at day of life 7 and 14. The online Neonatal BPD Outcome Estimator (NICHD Neonatal Research Network at https://neonatal.rti.org/index.cfm?fuseaction=BPDcalculator.start was then used to calculate the percent likelihood of developing each of five outcomes at 36 weeks PMA (none, mild, moderate, severe, death). The infants with X60% likelihood of having none or mild BPD as the outcome
