Timing of peritoneal dialysis initiation and mortality: analysis of the Canadian Organ Replacement Registry

Abstract

Background: Several observational studies of hemodialysis patients show an association between early dialysis therapy initiation and increased mortality. Few studies have examined this association among peritoneal dialysis patients.

Study design: Retrospective cohort study.

Setting & participants: A cohort of 8,047 incident peritoneal dialysis patients who started dialysis therapy in 2001-2009 and were treated in Canada.

Predictor: Estimated glomerular filtration rate (eGFR) at dialysis therapy initiation. Defined early, mid, and late starts as eGFR>10.5, 7.5-10.5, and <7.5mL/min/1.73m(2), respectively.

Outcomes: Time to death.

Measurements: Proportional piecewise exponential survival models to compare mortality (overall and early) for the 3 predictor groups.

Results: Between 2001 and 2009, the proportion of patients starting peritoneal dialysis therapy as early starts increased from 29% (95% CI, 26%-32%) to 44% (95% CI, 41%-47%). Compared with the late-start group, the overall mortality rate was not higher for the early- (adjusted HR, 1.08; 95% CI, 0.96-1.23) or mid-start (adjusted HR, 0.96; 95% CI, 0.86-1.09) groups. However, when examined yearly, patients in the early-start group were significantly more likely to die within the first year of dialysis therapy compared with those in the late-start group (adjusted HR, 1.38; 95% CI, 1.10-1.73), but not in subsequent years.

Limitations: Bias and residual confounding may have influenced the observed relationship between predictor and outcome.

Conclusions: Patients are initiating peritoneal dialysis therapy at increasingly higher eGFRs. Contrary to most observational studies assessing hemodialysis, the early initiation of peritoneal dialysis therapy, at eGFR>10.5mL/min/1.73m(2), is not associated with increased mortality.

Keywords: Estimated glomerular filtration rate (eGFR); mortality; peritoneal dialysis; retrospective cohort study; timing.