Shared decision making and dialysis choice: an observational longitudinal cohort study.

Abstract

Background: The 'Shared decision making and dialysis choice' intervention has been part of usual care at two hospitals in Denmark since 2018. The objective was to describe dialysis modality choice and outcomes for patients with kidney failure who received a shared decision making intervention.

Methods: Retrospective observational longitudinal cohort study design was used. Data were collected from 2018 to 2023 on 484 patients with kidney failure from one regional and one university hospital. The exposure was a shared decision making intervention for dialysis choice. The predictors were frailty, estimated glomerular filtration rate (eGFR), comorbidity, Body Mass Index (BMI), ethnicity, marital status and smoking. The outcomes were home-based dialysis, time, concordance, and death. Fisher's exact tests and Wilcoxon rank-sum tests assessed whether choice of dialysis modality differed significantly. Aalen-Johansen estimation assessed time from the shared decision making intervention to treatment initiation, concordance between chosen and initiated treatment, and mortality before treatment initiation. Logistic regression and Cox proportional hazards evaluated the patient characteristics predicting these three outcomes.

Results: After the intervention, 68% chose home-based dialysis, while 32% chose center-based dialysis. With significant differences, more patients aged ≤ 70 years, at the university hospital, and living with a partner chose home-based dialysis. Half of the patients initiated treatment within 11 months, and predictors for initiating dialysis later than 11 months were age ≥ 70 years and eGFR > 15 ml/min/1.73 m². 83% of the patients received the treatment chosen, and predictors for concordance were center-based dialysis, regional hospital, and very mild to mild frailty. 12% of the patients died before treatment initiation, predicted by very mild to severe frailty and BMI < 25 kg/m².

Conclusions: A high proportion of patients chose a home-based treatment after receiving the intervention and initiated their preferred dialysis choice. 50% of patients received the intervention 11 months before initiating dialysis, and few patients died before initiating dialysis. Routinely assessing frailty and BMI prior to intervention could possibly improve patient pathways. Complete follow-up for all patients was not ensured.