Effect of preoperative non-dialysis-dependent renal dysfunction on isolated aortic and mitral valve surgery: a propensity score analysis.

Abstract

Our objective was to examine whether preoperative non-dialysis-dependent renal dysfunction is associated with operative mortality or morbidity in isolated valve surgery. We reviewed consecutive patients undergoing isolated aortic (n = 2132) or mitral valve (n = 1664) surgery, between 1996 and 2009. Preoperative renal dysfunction was defined as preoperative estimated glomerular filtration rate < 60 mL/min without dialysis. Propensity score 1:1 matched samples were created, one for aortic (n = 626) and one for mitral (n = 526) valve surgery. The mean age was 70 ± 9 and 65 ± 10 years for the aortic and mitral groups, respectively. In the aortic cohort, patients with preoperative renal dysfunction had greater need for inotropes (39% vs 29%; P = .009), length of intensive care unit stay (27 vs 25 hours; P = .006), and duration of mechanical ventilation (8.2 vs 6.6 hours; P < .001). Operative mortality was 3.2% in the group with preoperative renal dysfunction and 2.2% in the group without preoperative renal dysfunction (P = .5). In the mitral cohort, patients with preoperative renal dysfunction had greater need for inotropes (47% vs 36%; P = .013), length of intensive care unit stay (40 vs 26 hours; P = .01), and duration of mechanical ventilation (7.2 vs 6.5 hours; P = .004). Operative mortality was 0% and 2.7% in the groups without and with preoperative renal dysfunction, respectively (P = .015). Preoperative renal dysfunction is associated with higher morbidity in both cohorts, and patients undergoing mitral valve surgery also experienced higher mortality. The impact of non-dialysis-dependent preoperative renal dysfunction appears to be more pronounced in patients undergoing mitral valve surgery, potentially owing to their relative intolerance to volume overload.