Patient-Reported Outcomes with High-Volume Hemodiafiltration Versus High-Flux Hemodialysis Across Clinical Subgroups

BACKGROUND: High-volume hemodiafiltration (HDF) improves survival and attenuates decline in selected patient-reported outcome (PRO) domains versus high-flux hemodialysis (HD). Whether benefits differ across clinically relevant subgroups remains uncertain. METHODS: We conducted a prespecified subgroup analysis of the multicenter CONVINCE randomized controlled trial comparing high-volume post-dilution HDF (≥23 L/session) with high-flux HD in adults receiving maintenance dialysis. Eight PROMIS domains were assessed quarterly over a median follow-up of 30 months. Linear mixed-effects models evaluated longitudinal PRO changes and treatment-by-subgroup interactions by prespecified baseline characteristics: age, sex, diabetes, cardiovascular disease, dialysis vintage, vascular access type, and residual urine volume. Joint-model sensitivity analyses of all eight longitudinal PROMIS trajectories and time to death assessed informative dropout due to death. RESULTS: Of 1360 randomized participants, 1211 with baseline and follow-up PRO assessments were included. PRO treatment effects were broadly consistent across prespecified clinical subgroups, without evidence of clinically meaningful heterogeneity. Models adjusted for baseline sociodemographic and clinical characteristics showed nominal treatment-by-subgroup interactions for dialysis vintage and physical function (P = 0.03), age and pain interference (P = 0.02), and dialysis vintage and sleep disturbance (P = 0.04). Joint-model sensitivity analyses accounting for informative dropout due to mortality reproduced the first two interactions and identified nominal interactions for vascular access type and sleep disturbance (P = 0.03) and cardiovascular disease status and cognitive function (P = 0.05). Interactions were modest in magnitude, exploratory, unadjusted for multiple comparisons, and not consistently observed across related PROMIS domains. CONCLUSIONS: This prespecified exploratory subgroup analysis found no convincing evidence of clinically meaningful treatment-effect heterogeneity across baseline characteristics examined. Nominal interactions were unadjusted for multiple comparisons and should be regarded as hypothesis-generating. Findings should not guide selection or exclusion of patients for high-volume HDF based on patient-reported outcomes.