Propensity score matched mortality comparisons of peritoneal and in-centre haemodialysis: systematic review and meta-analysis.
Pooling the best-matched observational studies of more than 113,000 patients, peritoneal dialysis and in-centre haemodialysis were associated with essentially equivalent survival, suggesting earlier reported differences mostly reflected patient selection rather than the treatments themselves.
Background
Comparing HD and PD survival from observational data is difficult because of substantial residual confounding from baseline imbalances between patients selected for each modality. Propensity-score matching (PSM) reduces this selection bias beyond conventional adjustment.
Study design
Systematic review and random-effects meta-analysis of PSM-based cohort studies identified in PubMed, Cochrane and EMBASE from inception through November 2018, with meta-regression to explain between-study variation. Seventeen cohort studies covering 113,578 PSM incident dialysis patients (cohort periods 1993-2014) were included.
Key findings
The pooled hazard ratio for PD versus HD was 1.06 (95% CI 0.99-1.14). Risks were similar across regions (Europe 1.04, Asia 1.14, Americas 0.98) and in patients with diabetes (1.09) and without diabetes (0.99). Heterogeneity was high (I-squared 87%) and was largely explained by differences in cohort period, study type and country (R-squared 90.6%).
Clinical implications
PD and in-centre HD appear to offer equivalent survival; reported differences largely reflect factors unrelated to clinical efficacy. This supports shared decision-making centred on patient lifestyle, preference and access rather than an assumed survival difference between modalities.
Category
Research
Source
Nephrology Dialysis Transplantation
More from ASNRT News
Browse the latest news, society announcements, KDIGO guideline updates, and AJNT issue releases on the ASNRT newsroom.