Residual urine output and postoperative mortality in maintenance hemodialysis patients.

BACKGROUND The relationship between residual urine output and postoperative survival in maintenance hemodialysis patients is unknown. OBJECTIVE To explore the relationship between amount of urine before surgery and postoperative mortality and differences between postoperative nonanuria and anuria in maintenance hemodialysis patients. METHODS A total of 109 maintenance hemodialysis patients underwent major operations. Anuria was defined as urine output <30 mL in the 8 hours before the first session of postoperative dialysis. Propensity scores for postoperative anuria were developed. RESULTS Postoperative residual urine output was 159.2 mL/8 h (SD, 115.1) in 33 patients; 76 patients were anuric. Preoperative residual urine output and adequate perioperative blood transfusion were positively related to postoperative urine output. Propensity-adjusted 30-day mortality was associated with postoperative anuria (odds ratio [OR], 4.56; 95% confidence interval [CI], 1.16-17.96; P = .03), prior stroke (OR, 4.46; 95% CI, 1.43-13.89; P = .01) and higher disease severity (OR, 1.10; 95% CI, 1.00-1.21; P = .049) at the first postoperative dialysis. OR of 30-day mortality was 5.38 for nonanuria to anuria vs nonanuria to nonanuria (P = .03) and 5.13 for preoperative anuria vs nonanuria to nonanuria (P = .01). By Kaplan-Meier analysis, 30-day mortality differed significantly among patients for nonanuria to nonanuria, anuria, and nonanuria to anuria (log rank, P = .045). CONCLUSION Patients with preoperative nonanuria and postoperative anuria had higher mortality than did patients with no anuria before and after surgery and patients with anuria before surgery. Postoperative residual urine output is an important surrogate marker for disease severity.