verall outcomes of neonatal open-heart surgery have been dramatically improved over the past 2 decades owing to refinements of preoperative, intraoperative and postoperative management. The current early mortality rate of patients with transposition of the great arteries (TGA) with intact atrial septum (TGA-IAS) and TGA with ventricular septal defect (TGA-VSD) undergoing arterial switch operation (ASO) is approximately 2-7%, 1,2 a considerable improvement from the mortality rate of about 15% in earlier eras. 3 Interest has therefore been shifted from reducing operative mortality to facilitating quicker recovery and reducing intensive care unit (ICU) stay and hospital stay by applying an early extubation policy and/or fast-track cardiac surgery pathways. [4][5][6] However, stress response to cardiopulmonary bypass (CPB) is much greater in neonates than in older children, 7 and postoperative recovery in neonates can therefore often be compromised by CPB-induced organ dysfunction.Modified ultrafiltration (MUF), which was introduced by Naik et al in 1991, has become an essential perfusion strategy to minimize CPB-induced adverse effects in pediatric patients undergoing open-heart surgery. 8 A substantial body of evidence showed that MUF improves cardiac, pulmonary and cerebral functions 9-12 and decreases postoperative bleeding and blood transfusion requirement 13 by reduction of body water accumulation, reversal of hemodilution, and modulation of systemic inflammatory mediators. 9,14 The beneficial effects of MUF are thought to be greatest in neonates who have the largest body weight/CPB circuit volume mismatch and are therefore most vulnerable to exposure to hypothermia and crystalloid hemodilution; however, the effect of MUF on clinical outcomes in neonates undergoing open-heart surgery has not been clarified in previous studies. 13,15,16 We hypothesized that MUF accelerates the recovery in neonates undergoing open-heart surgery by early recovery of hemodynamics and pulmonary function and by attenuation of capillary leakage, thereby contributing to our 'fast-track' strategy. To test our hypothesis, we analyzed patients with TGA-IAS and TGA-VSD who underwent ASO with or without MUF because of relative homogeneity of the patients' characteristics, standardized surgical procedure, low operative mortality, and favorable early and long-term hemodynamic results, 17,18 providing a better picture of impacts of MUF on early clinical outcomes in neonatal patients.
MethodsWe conducted a retrospective study of neonates who underwent ASO from April 1998 to September 2006 at Okayama University Hospital. The Institutional Review