Abstract
BACKGROUND AND OBJECTIVES: Automated peritoneal dialysis (APD) is increasingly used in comparison with continuous ambulatory peritoneal dialysis (CAPD). Although APD is expected to improve survival, convincing evidence of major advantages is lacking. The objective was to investigate whether overall mortality and technique failure of incident dialysis patients treated with APD are different from those treated with CAPD. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: Patients on APD or CAPD at 3 mo after start of dialysis were selected from a prospective multicenter cohort study in incident dialysis patients (NECOSAD). Overall mortality was studied with an intention-to-treat design; the event was death. Technique failure was studied with an as-treated design; the event was a switch of dialysis modality. Hazard ratios (HRs) were calculated with a follow-up of 5 yr. The HRs were adjusted for gender, age, primary kidney disease, comorbidity, residual GFR, urine production and plasma albumin at 3 mo after inclusion. RESULTS: Eighty-seven APD and 562 CAPD patients were included. In the intention-to-treat analysis 154 CAPD and 21 APD patients died. The crude HR for overall mortality was 0.98 (95% CI: 0.62-1.54), the adjusted HR was 1.09. In the as-treated analysis 238 CAPD and 34 APD patients switched therapy, whereas 91 CAPD and 7 APD patients died. The crude HR for technique failure was 0.92 (95% CI: 0.64-1.31) and did not change after adjustment. CONCLUSIONS: No difference was found in overall mortality and technique failure for APD compared with CAPD in incident dialysis patients.