Predictors and early and late outcomes of dialysis-dependent patients in contemporary cardiac surgery

Objectives: The aim of the study was to investigate the incidence and predictors of renal failure requiring dialysis (RF-D) in a contemporary cohort of patients undergoing cardiac surgery. The authors also analyzed early and late outcome of patients with this complication. Design: A retrospective study of consecutive patients undergoing cardiac surgery using a computerized database based on the New York State Department of Health registry. Data collection was performed prospectively. Setting: A university hospital (single institution). Participants: Six thousand four hundred forty-nine patients who underwent cardiac surgery between January 1998 and December 2006 including isolated coronary artery bypass graft (CABG) surgery (n = 2,819, 44%), single- or multiple-valve surgery (n = 1,378, 21%), combined valve and CABG procedures (n = 1,032, 16%), and surgery involving the ascending aorta or the aortic arch (n = 1,220, 19%). Interventions: None. Measurements and Main Results: The incidence of RF-D was 2.2% (n = 139). The incidence per type of procedure was as follows: CABG surgery (0.8%), valve/CABG surgery (2.7%), valve surgery (2.9%), and aortic surgery (4%) (p = 0.001). Multivariate analysis revealed preoperative renal dysfunction (odds ratio [OR] = 5.5), hemodynamic instability (OR = 5.2), diabetes (OR = 2.6), aortic surgery (OR = 2.2), congestive heart failure (CHF) (OR = 2.1), peripheral vascular disease (PVD) (OR = 1.9), and reoperation (OR = 1.8) as independent predictors of RF-D. The hospital mortality after RF-D was 36.7% (n = 51) compared with 2.9% (n = 180) in the control group (p < 0.001). Long-term survival after RF-D was significantly decreased (1-year and 5-year survival 48.5% +/- 6.1% and 28.7% +/- 7.2% v 94.5% +/- 0.3% and 83.5% +/- 0.6% in the control group, p < 0.001). Hypertension, CHF, and PVD were independent predictors of late mortality. Conclusion: The authors observed an increase in the overall incidence of RF-D compared with previous studies, probably related to an increased prevalence of patients undergoing more complex procedures with a worsening risk profile. Postoperative RF-D was not only associated with increased hospital mortality and morbidity, but also with a significant reduction of long-term survival in discharged patients. Seven independent predictors of RF-D were identified. Future research efforts should focus on a more precise identification of patients at risk and the development of new treatment modalities, which would potentially prevent the occurrence of this complication. (C) 2008 Elsevier Inc. All rights reserved.