Comparative mortality of hemodialysis and peritoneal dialysis in Canada

Background. Comparisons of mortality rates in patients on hemodialysis vers us those on peritoneal dialysis have been inconsistent. We hypothesized tha t comorbidity has an important effect on differential survival in these two groups of patients. Methods. Eight hundred twenty-two consecutive patients at 11 Canadian insti tutions with irreversible renal failure had an extensive assessment of como rbid illness collected prospectively, immediately prior to starting dialysi s therapy. The cohort was assembled between March 1993 and November 1994; v ital status was ascertained as of January 1, 1998. Results. The mean follow-up was 24 months. Thirty-four percent of patients at baseline, 50% at three months, and 51% at six months used peritoneal dia lysis. Values for a previously validated comorbidity score were higher for patients on hemodialysis at baseline (4.0 vs. 3.1, P < 0.001), three months (3.7 vs. 3.2, P = 0.001), and six months (3.6 vs. 3.2 P = 0.005). The over all mortality was 41%. The unadjusted peritoneal dialysis/hemodialysis mort ality hazard ratios were 0.65 (95% CI, 0.51 to 0.83, P = 0.0005). 0.84 (95% CI, 0.66 to 1.06, P = NS). and 0.83 (95% CI, 0.64 to 1.08, P = NS) based o n the modality of dialysis in use at baseline, three months. and six months , respectively. When adjusted for age, sex, diabetes, cardiac failure, myoc ardial infarction, peripheral vascular disease, malignancy, and acuity of r enal failure, the corresponding hazard ratios were 0.79 (95% CI, 0.62 to 1. 01, P = NS), 1.00 (95% CI, 0.78 to 1.28, P = NS), and 0.95 (95% CI, 0.73 to 1.24, P = NS). Adjustment for a previously validated comorbidity score res ulted in hazard ratios of 0.74 (95% CI, 0.58 to 0.94, P = 0.01), 0.94 (95% CI. 0.74 to 1.19, P = NS), and 0.88 (95% CI, 0.68 to 1.13, P = NS) at basel ine, three months, and six months. There was no survival advantage for eith er modality in any of the major subgroups defined by age, sex, or diabetic status. Conclusions. The apparent survival advantage of peritoneal dialysis in Cana da is due to lower comorbidity and a lower burden of acute onset end-stage renal disease at the inception of dialysis therapy. Hemodialysis and perito neal dialysis, as practiced in Canada in the 1990s. are associated with sim ilar overall survival rates.