Impact of time-varying center volume on technique failure and mortality in peritoneal dialysis.

Abstract

Most studies on volume-outcome association used the number of patients at a particular period as the independent variable. However, peritoneal dialysis (PD) is a chronic treatment, and center volume usually changes over a patient's treatment period. Accordingly, this study used the time-varying center volume to explore the volume-outcome association in PD. We conducted a nationwide population-based retrospective cohort study, which included patients who began chronic PD between 2001 and 2010. The risk factors of 5-year technique failure and mortality were analyzed using cause-specific and subdistribution hazard models, respectively. The annual number of patients initiating PD in each patient's treatment center was modeled as a time-varying variable with four categories. We included 9071 patients who started PD in 100 centers where the number of incident patients ranged from 1 to 107 patients per year (median, 25; interquartile range, 13-42). The estimated 5-year patient and technique survival rates were 64.7% and 66.6%, respectively. Being treated in centers in the largest volume category (the number of incident PD patients ≥43 per year) was associated with significantly lower cause-specific and cumulative hazards for technique failure. No association was found between facility volume and hazards of mortality. Receiving PD in high-volume facilities was associated with a lower risk in technique failure. No association was found between facility volume and mortality risk.