[Congenital diaphragmatic hernia: an analysis of the results and prognostic factors prior to the development of an ECMO program].

Background: Some neonates with congenital diaphragmatic hernia (CDH) and persistent pulmonary hypertension are not adequately oxygenated with conventional treatment. The extracorporeal membrane oxygenation (ECMO) has been successful in some of them as an alternative in their management.

Patients And Methods: We studied the charts of 47 neonates with CDH, symptomatic within 24 hours of birth, treated in our institution during the last seven years (1987-1994). In all of them, conventional ventilation and hemodynamic support was used. In 12 patients high frequency ventilation (HFV) was used and two survived. In all patients we analyzed the following ventilatory and gasometric parameters: Oxygenation index (OI)*, ventilatory index (VI)** and postductal PCO2. In 15 neonates who did not survive, a necropsy was performed and a morphometric parameter, pulmonary index (PI)*** was studied.

Results: The overall survival was 60%. VI and OI showed significant differences (p < 0.001) between survivors and non-survivors with values of 460.9 +/- 303 vs 1532 +/- 500.6, respectively for VI and 10.3 +/- 5.7 vs 46.2 +/- 37.8, respectively for IO. There were no significant differences in postductal PCO2. Mean PI in the 15 non-survivors was 0.0072 +/- 0.002 (normal > 0.015). Regression coefficients of PI with OI or VI were not significant. Neonates with VI < 1000 and OI < 40 survived. All patients with VI > 1000 and OI > 40 died. Some babies with VI > 1000 and OI < 40 (21.6%) survived.

Conclusions: In our experience, the use of HFV did not improve the prognosis of these patients, but we believe that the use of ECMO in those patients with VI > 1000, and overall, patients with VI > 1000 and OI < 40 would improve the survival rates of this congenital malformation. *QI = FiO2 x MAP/PO2 postductal x 100. (MAP = Median airway pressure). **VI = VR x MAP (VR = Ventilatory rate). ***PI = Pulmonary weight/Body weight.