13 Years Experience With Congenital Diaphragmatic Hernia (CDH) in a Large Inborn Population. 1521

Few institutions have sufficient deliveries to accurately evaluate incidence, outcome, and outcome variables predictive for neonates with CDH. Parkland Memorial Hospital averages >13,000 live births (LB)/year; thus we performed a longitudinal population-based study on 180,643 LB over 13yr(1983-1995), reviewing the charts of all high-risk neonates with CDH, defined as those symptomatic within the first six hours of life, and recording prenatal data, presenting symptoms, birth weight (BW), gestational age, sex, Apgar scores, presence and type of other anomalies, side of hernia, preoperative arterial blood gases, time from birth to surgery, and use of ECMO. 43 neonates (1/4,200 LB) had CDH; 16% were right-sided, 42% had other anomalies, and overall survival was 51%. While 80% of neonates with isolated CDH survived, 89% with CDH and other anomalies died. 11 neonates did not undergo surgery; 10 of whom had anomalies incompatible with life, and one with severe pulmonary hypoplasia. Surgical correction was performed in 32 high-risk neonates; 69% survived and only 6 (19%) received ECMO. Those who died had lower BW (2.3±0.2 vs 3.1±0.1 kg; p=.004, t-test), were less mature (36.3±0.8 vs 38.3±0.5 wks EGA; p=0.15), more acidotic(7.22±0.06 vs 7.38±0.02; p=0.0001), and had more severe respiratory compromise, i.e., higher ventilation index (VI), oxygenation index, and AaDO₂. Using the “best” preoperative PaCO₂ and VI with “Bohn's quadrant analysis”, neonates with PaCO₂ <40 mmHg and VI <1000 had 92% survival. 1/5 in Bohn's 100% mortality quadrant survived. Survival was 60% and 67% when PaCO₂ was>40 mmHg with low VI and PaCO₂ <40 mmHg with high VI, respectively. Median (25 and 75 percentiles) time to surgery was unchanged during the study period: 6.65h (5.2, 10) for survivors and 4.95h (2.6,5.0) for nonsurvivors; 88% had surgery <24h of life. In this inborn population-based study, the best predictors of survival for CDH were the presence or absence of other anomalies and the “best” preoperative PaCO₂-VI using“Bohn's quadrant analysis”. ECMO was infrequently needed, suggesting its use might benefit neonates without other anomalies who are predicted to fail conventional management by Bohn's criteria.