Early Prediction of Chronic Oxygen Dependency by Lung Function Test Results V. Kavvadia, MD, A. Greenough, FRCPCH,* and G. Dimitriou, MD Summary. Chronic oxygen dependency (COD) is a common sequela to very premature birth. Steroid therapy may reduce COD if given within the first 2 weeks, but has important side effects. It is, therefore, crucial to identify an accurate predictor of COD and hence only expose high-risk infants to intervention therapy. The aim of this study was to determine if, within 48 hr of birth, abnormal lung function predicted COD and whether such results performed better than readily available clinical data. Results from 100 consecutive, very low birth-weight infants, median gestation age 28 weeks (range, 24–33), who were ventilated within 6 hr of birth and survived beyond 36 weeks postconceptional age (PCA), were analyzed. Lung volume was assessed by measurement of functional residual capacity (FRC) using a helium gas dilution technique, and compliance was measured using either a passive inflation or an occlusion technique. The maximum peak inflating pressure and inspired oxygen concentration within the first 48 hr were recorded. The infants who remained oxygen-dependent beyond 28 days (n = 58) and 36 weeks PCA (n = 24) differed from the rest in being more immature (P < 0.001), more had a patent ductus arteriosus, and they had both a lower median lung volume (P < 0.001) and lower compliance (P < 0.01) on day 2. An FRC <19 mL/kg and a low gestational age were the most accurate predictors of COD at 28 days. An FRC <19 mL/kg on day 2 remained the best predictor of COD beyond 28 days if only the 50 infants whose gestational age was 28 weeks were considered. We conclude that demonstration of a low lung volume in the first 48 hr helps to identify infants who might benefit from therapy aimed at preventing COD. Pediatr Pulmonol. 2000; 29:19– 26. © 2000 Wiley-Liss, Inc. Key words: chronic lung disease; prematurity; lung volume; compliance; infant pulmonary function; oxygen therapy; neonates; dexamethasone. INTRODUCTION Approximately 1–2% of newborn babies develop re- spiratory distress requiring mechanical ventilation. A sig- nificant proportion develop chronic oxygen dependency (COD). Affected infants have protracted neonatal inten- sive care unit admissions and, once discharged, suffer ongoing respiratory morbidity. Steroids are routinely given to infants at risk of COD, but a recent meta- analysis 1 demonstrated that only if steroid therapy is started in the first 2 weeks does it reduce COD and mortality. Unfortunately, systemic steroids have many side-effects and there is increasing concern 2 regarding their effect on lung growth 3 and neurological outcome. 4 As a consequence, it has been recommended that if early dexamethasone is to be used for the prevention of COD, the critical time to initiate the therapy must be further defined. 1 Crucial to that process is the development of an easily performed and accurate test which will identify infants at high risk of COD. Infants who develop COD have abnormal lung func- tion in the first week of life, but there is controversy regarding the precise abnormality. A high pulmonary re- sistance, approximately 90% higher than in “controls,” has been reported in those destined to develop COD. 5 Others 6 have demonstrated a low dynamic pulmonary compliance in infants who subsequently develop COD. Infants with low lung volume also appear at risk of de- veloping COD, but the measurements were made at 14 days. 7 The predictive value of a low lung volume in the first days after birth has not been compared in a high-risk Children Nationwide Regional Neonatal Intensive Care Centre, King’s College Hospital, London, UK. Grant sponsor: South Thames Regional Health Authority’s Research and Development Directorate; Grant sponsor: Children Nationwide/ Nestle Research Fellowship; Grant sponsor: Children Nationwide Medical Research Fund; Grant sponsor: Royal Society. *Correspondence to: A. Greenough, F.R.C.P.C.H., Children Nation- wide Regional Neonatal Intensive Care Centre, King’s College Hos- pital, London SE5 9RS, UK. Received 23 April 1999; Accepted 20 August 1999. Pediatric Pulmonology 29:19–26 (2000) © 2000 Wiley-Liss, Inc.