Patient Preferences for Longer or More Frequent In-Center Hemodialysis Regimens: A Multicenter Discrete Choice Study

Rationale & Objective: Longer and more frequent hemodialysis sessions are associated with both benefits and harms. However, their relative importance to patients and how they influence acceptability for patients have not been quantified.

Study Design: Discrete-choice experiment in which a scenario followed by 12 treatment choice sets were presented to patients in conjunction with varying information about the clinical impact of the treatments offered.

Setting & Participants: Patients with kidney failure treated with maintenance dialysis for≥1 year in 5 UK kidney centers.

Predictors: Length and frequency of hemodialysis sessions and their prior reported associations with survival, quality of life, need for fluid restriction, hospitalization, and vascular access complications.

Outcome: Selection of longer (4.5 hours) or more frequent (4 sessions per week) hemodialysis regimens versus remaining on 3 sessions per week with session lengths of 4 hours.

Analytical Approach: Multinomial mixed effects logistic regression estimating the relative influence of different levels of the predictors on the selection of longer and more frequent dialysis, controlling for patient demographic characteristics.

Results: Among 183 prevalent in-center hemodialysis patients (mean age of 63.7 years, mean dialysis vintage of 4.7 years), 38.3% (70 of 183) always chose to remain on regimens of 3 sessions per week with session duration of 4 hours. Depicted associations of increasing survival and quality of life, reduced need for fluid restriction, and avoiding additional access complications were all significantly associated with choosing longer or more frequent treatment regimens. Younger age, fatigue, previous experience of vascular access complications, absence of heart failure, and shorter travel time to dialysis centers were associated with preference for 4 sessions per week. Patients expressed willingness to trade up to 2 years of life to avoid regimens of 4 sessions per week or access complications. After applying estimated treatment benefits and harms from existing literature, the fully adjusted model revealed that 27.1% would choose longer regimens delivered 3 times per week and 34.3% would choose 4 hours 4 times per week. Analogous estimates for younger fatigued patients living near their unit were 23.5% and 62.5%, respectively.

Limitations: Estimates were based on stated preferences rather than observed behaviors. Predicted acceptance of regimens was derived from data on treatment benefits and harms largely sourced from observational studies.

Conclusions: Predicted acceptance of longer and more frequent hemodialysis regimens substantially exceeds their use in current clinical practice. These findings underscore the need for robust data on clinical effectiveness of these more intensive regimens and more extensive consideration of patient choice in the selection of dialysis regimens.