Aims: (1) To determine incidence and reasons for transfer from maintenance haemodialysis (HD) to peritoneal dialysis (PD); (2) To compare mortality of HD patients transferred to PD with those who initiated and remained on PD.
Methods: A 1:2 matched cohort of ANZDATA included patients dialyzing over 3 months dividing into 2 groups. Group A (HD to PD) was compared to group B (initiated and remained on PD) and matched PD group (Group C). We compared mortality by total time on dialysis regardless of modality.
Results: Of 20 882 patients, there were 911 in group A. The transfer rate from HD to PD was 5%, 6.7% and 7.4% at 1,4 and 8 years, respectively. Median time before switching of patients in Group A was 5.9 (4.0-10.9) months. The commonest reported reason for transfer was patient preference (63.8%). Mortality was significantly worse in Group A if dialysis vintage was ≤6 months and from 12-24 months (P < 0.05), whereas there was no difference in mortality if dialysis vintage was 6-12 months or ≥ 24 months (P = 0.073 and P = 0.153, respectively). Overall, mortality of patients in group A was higher than that in group B after adjusting for age, race, chronic lung disease, peripheral vascular disease, cerebrovascular disease, diabetes, haemoglobin and phosphate (hazard ratio = 1.335, 95% confidence interval = 1.172-1.520).
Conclusion: Haemodialysis patients transferred to PD had higher mortality than those initiated on PD, strengthening the case for PD first policy.
Keywords: end-stage kidney disease; haemodialysis; peritoneal dialysis; transfer.
© 2018 Asian Pacific Society of Nephrology.