Abstract
Between 4 and 16% of extremely premature infants have late pulmonary hypertension (PH) (onset >30 days of life), and infants with PH have a higher risk of tracheostomy and death. Atrial septal defects (ASD) increase pulmonary blood flow and may promote PH in at-risk infants. The objective of this study was to determine if infants with ASD develop PH sooner than those without ASD. Infants who were born at < 32 weeks' gestation, with an echocardiogram on day of life > 30, and without congenital anomalies were included. Infants with and without ASD were evaluated for the time to PH diagnosis, defined as the day of the first echocardiogram that showed PH. A multivariable model with ASD and significant variables on PH and a Cox proportional hazard model evaluating time to PH was determined. Of the 334 infants with echocardiograms, 57 had an ASD and 26% of these developed PH vs. 12% without ASD (p = 0.006). Infants with PH had lower gestational age (25.2 vs. 26.2 weeks, p = 0.005), smaller birthweight (699 vs. 816 gm, p = 0.001), and more prematurity complications than infants without PH. More PH infants had maternal African-American race (63.9 vs. 36.1%), right ventricular dysfunction (23.9 vs. 3.2%, p < 0.001), right ventricular dilation (52.1 vs. 8.6%, p < 0.001), or right ventricular hypertrophy (51.2 vs. 10.1%, p < 0.001), than infants without PH. At 150 days of life, 78.1% (95% CI 64.6-86.9%) of infants with ASD survived without PH, compared with 90.9% (95% CI 86.7-93.8%) of infants without ASD, and the unadjusted hazard for development of PH for infants with ASD was 2.37 (95% CI 1.29-4.36). When significant clinical variables were controlled, infants with ASD had a 2.44-fold (95% CI 1.27-4.68) increase in PH, compared with infants without ASD. Most PH in infants with or without ASD was diagnosed by day of life 150, but infants with ASD had an over 2-fold increased hazard for PH during their neonatal hospitalization. Premature infants with ASD should be followed closely for PH development and further studies to investigate the optimal timing of closure are needed.