Symptom Burden before and after Dialysis Initiation in Older Patients

Abstract

Background and objectives: For older patients with kidney failure, lowering symptom burden may be more important than prolonging life. Dialysis initiation may affect individual kidney failure-related symptoms differently, but the change in symptoms before and after start of dialysis has not been studied. Therefore, we investigated the course of total and individual symptom number and burden before and after starting dialysis in older patients.

Design, setting, participants, & measurements: The European Quality (EQUAL) study is an ongoing, prospective, multicenter study in patients ≥65 years with an incident eGFR ≤20 ml/min per 1.73 m². Using the dialysis symptom index (DSI), 30 symptoms were assessed every 3-6 months between 2012 and 2021. Scores for symptom number range from zero to 30 and, for burden, from zero to 150, with higher scores indicating more severity. Using mixed effects models, we studied symptoms during the year preceding and the year after dialysis initiation.

Results: We included 456 incident patients on dialysis who filled out at least one DSI during the year before or after dialysis. At dialysis initiation, mean (SD) participant age was 76 (6) years, 75% were men, mean (SD) eGFR was 8 (3) ml/min per 1.73 m², 44% had diabetes, and 46% had cardiovascular disease. In the year before dialysis initiation, symptom number increased +3.6 (95% confidence interval [95% CI], +2.5 to +4.6) and symptom burden increased +13.3 (95% CI, +9.5 to +17.0). In the year after, symptom number changed -0.9 (95% CI, -3.4 to +1.5) and burden decreased -5.9 (95% CI, -14.9 to -3.0). At dialysis initiation, "fatigue," "decreased interest in sex," and "difficulty becoming sexually aroused" had the highest prevalence of 81%, 69%, and 68%, respectively, with a burden of 2.7, 2.4, and 2.3, respectively. "Fatigue" somewhat improved after dialysis initiation, whereas the prevalence and burden of sexual symptoms further increased.

Conclusions: Symptom burden worsened considerably before and stabilized after dialysis initiation. "Fatigue," "decreased interest in sex," and "difficulty becoming sexually aroused" were considered most burdensome, of which only "fatigue" somewhat improved after dialysis initiation.

Keywords: chronic kidney disease; dialysis; elderly; end stage kidney disease; epidemiology and outcomes.

Graphical abstract

Figure 1.

Symptom number and symptom burden worsened considerably in the year before and stabilized in the year after start of dialysis in 456 older patients. These results represent the change in total symptom number and burden during the year preceding and after dialysis initiation. To obtain these results, linear mixed models were used in which time (days before or after start of dialysis) was modeled in a three-knot restricted cubic spline function with 95% confidence intervals (95% CIs) to allow for more flexibility. The knots were chosen at the start of dialysis initiation, 6 months before dialysis initiation, and 6 months after start of dialysis initiation. A random intercept and slope for time were used to account for repeated measurements, allowing the trajectory before and after the discontinuity to vary between individuals. DSI, dialysis symptom index.

Figure 2.

At dialysis initiation, “fatigue,” “decreased interest in sex,” and “difficulty becoming sexually aroused” had the highest mean symptom prevalence and burden (x axis) according to the five-point Likert scale (legend) of 30 kidney failure–related symptoms in 278 older patients during the 30 days before and after start of dialysis.

Figure 3.

“Fatigue,” “decreased interest in sex,” and “difficulty becoming sexually aroused” were the most prevalent and burdensome of all 30 kidney failure-related symptoms during the year before and after starting dialysis. Prevalence (dotted line, right y axis) and burden (solid line, left y axis) of 30 kidney failure–related symptoms in the year before and after start of dialysis in 456 older patients, ordered by their nine corresponding symptom systems.