Aggressive Surgical Management of Congenital Diaphragmatic Hernia: Worth the Effort?: A Multicenter, Prospective, Cohort Study.

Abstract

The objectives of this study were (i) to evaluate infants with congenital diaphragmatic hernia (CDH) that do not undergo repair, (ii) to identify nonrepair rate by institution, and (iii) to compare institutional outcomes based on nonrepair rate. Approximately 20% of infants with CDH go unrepaired and the threshold to offer surgical repair is variable. Data were abstracted from a multicenter, prospectively collected database. Standard clinical variables, including repair (or nonrepair), and outcome were analyzed. Institutions were grouped based on volume and rate of nonrepair. Preoperative mortality predictors were identified using logistic regression, expected mortality for each center was calculated, and observed /expected (O/E) ratios were computed for center groups and compared by Kruskal-Wallis ANOVA. A total of 3965 infants with CDH were identified and 691 infants (17.5%) were not repaired. Nonrepaired patients had lower Apgar scores (P < 0.05) and increased incidence of anomalies (P < 0.0001). Low-volume centers ("Lo", n=44 total, < 10 CDH pts/yr) and high-volume centers ("Hi", n = 21) had median nonrepair rates of 19.8% (range 0%-66.7%) and 16.7% (5.1%-38.5%), respectively. High-volume centers were further dichotomized by rate of nonrepair (HiLo = 5.1-16.7% and HiHi = 17.6-38.5%), leaving 3 groups: HiLo, HiHi, and Lo. Predictors of mortality were lower birth weight, lower Apgar scores, prenatal diagnosis, and presence of congenital anomalies. O/E ratios for mortality in the HiLo, HiHi, and Lo groups were 0.81, 0.94, and 1.21, respectively (P < 0.0001). For every 100 CDH patients, HiLo centers have 2.73 (2.4-3.1, 95% confidence interval) survivors beyond expectation. There are significant differences between repaired and nonrepaired CDH infants and significant center variation in rate of nonrepair exists. Aggressive surgical management, leading to a low rate of nonrepair, is associated with improved risk-adjusted mortality.