VENOVENOUS EXTRACORPOREAL MEMBRANE-OXYGENATION AFFECTS RENAL-FUNCTION

Objective. We evaluated the effect of venovenous extracorporeal membra ne oxygenation (ECMO) on renal function and fluid balance in neonates with severe respiratory failure. Design. We retrospectively reviewed t he charts of 30 consecutive patients who met criteria for treatment wi th ECMO. Twelve were managed without ECMO (comparison group) and 18 we re treated with venovenous ECMO (treatment group). Setting. The study was conducted in a single level III neonatal intensive care unit in a regional children's hospital accepting medical and surgical neonatal t ransfers. Our hospital does not have an inborn service. Patients. Neon ates were included if their gestational age was more than 34 weeks, th ey weighed more than 2 kg, and their respiratory failure was severe en ough to warrant consideration of ECMO as a mode of support. All the ne onates in this study were treated with high-frequency ventilation befo re being considered for ECMO; none were treated with nitric oxide. Cri teria used to determine whether a neonate was a candidate for ECMO inc luded: (1) alveolar-arterial oxygen difference greater than 60 kPa (61 0 torr) for 8 hours; (2) alveolar-arterial oxygen difference greater t han 59 kPa (605 torr) and a peak airway pressure greater than 3.7 kPa (38 cm H2O) for 4 hours; (3) oxygenation index greater than 40 on thre e of five postductal blood gases obtained at least 30 minutes apart an d unstable patient condition; or (4) refractory, severe respiratory fa ilure with sudden decompensation (partial pressure of arterial oxygen 3.4 kPa or lower, 35 torr) despite maximal medical management for 2 ho urs. We did not include patients with congenital diaphragmatic hernia. Main results. There were no differences between the groups in gestati onal age, birth weight, age at admission, gender, or diagnoses. Over t he course of the 108 hours reviewed for each case, neonates treated wi th ECMO had higher positive fluid balance (P < .001), lower urine flow rates (P < .01), and higher blood urea nitrogen (P < .01) and creatin ine (P < .01) levels than neonates managed without ECMO. There were no differences in mean blood pressure, protein intake, serum albumin or use of diuretic therapy that might explain the differences between the groups. Conclusion. We conclude that venovenous ECMO is associated wi th transient impairment in renal function and marked fluid retention.