Background: While the number of patients returning to dialysis after graft failure (GF) is increasing steadily, the impact of a failed kidney transplant on mortality among dialysis patients has not been studied well.
Methods: Data from the Canadian Organ Replacement Register were utilized to examine the outcomes of an incident cohort of patients (n = 25,632) initiating renal replacement therapy (RRT) between 1990 and 1998. Cox regression was used to compare covariate-adjusted mortality among five RRT categories: transplant-naive dialysis, cadaveric primary renal transplant, living-donor primary renal transplant, post-GF dialysis and retransplant. RRT category-specific hazard ratios (HR) were estimated using Cox regression and adjusting for age, sex, race, calendar period, primary renal diagnosis and comorbid conditions.
Results: Mortality among post-GF dialysis patients was approximately equal to that of transplant-naive patients (HR = 0.90; P = 0.30) while the HR for retransplanted patients was significantly decreased, relative to the transplant-naive group (HR = 0.35; P<0.01). Diabetes was found to be a significantly (P<0.01) stronger mortality risk factor among post-GF dialysis patients (HR = 3.71) compared with the transplant-naive group (HR = 1.73). In the post-GF group, cardiovascular disease (HR = 1.66) and 'other serious illness' (HR = 2.07) were found to be much stronger risk factors for mortality than in the transplant-naive group (HR = 1.33 and 1.43, respectively), although the differences failed to reach statistical significance.
Conclusions: These results suggest that transplant-naive and post-GF dialysis patients have equivalent mortality risk and that mortality is significantly reduced upon retransplantation. In addition, the results highlight the importance of diabetes and, possibly, comorbid conditions as potential modifiable risk factors in the management of post-GF dialysis patients.