ABSTRACT
INTRODUCTION: Modifiable hemodialysis treatment parameters may impact patient reported outcomes, including recovery time. Answers to the recovery question may predict the impact of treatment parameters on clinical outcomes and health related quality of life. However, the reliability of answers to the recovery question after consecutive and nonconsecutive dialysis treatments in diverse populations has not been established. OBJECTIVE: To assess the reliability of this instrument and to determine if recovery time was associated with modifiable dialysis parameters, we conducted a quality assurance project in which we asked, "How long did it take you to recover from your last dialysis session?" after consecutive and nonconsecutive treatments. METHODS: We asked patients the recovery question ≤ seven times. We computed polychoric correlations to assess within patient correlations. We used random intercept ordinal logistic regression models to test for associations of recovery time with patient variables. RESULTS: We obtained answers to the recovery question in association with 1572 treatments in 364 patients. Recovery time was <2 hours in 52.1%; 2 to 7 hours in 20.9%; and >7 hours in 27.0% of treatments. Polychoric correlations demonstrated highly reliable responses within individual patients between consecutive and nonconsecutive treatments. Prolonged recovery was associated with a dialysate potassium of 1 vs. 2 mEq/L (odds ratio [OR] 2.25 {95% confidence interval [CI] 1.43-3.55}) and 1 vs. 3 mEq/L (OR 1.88 [95% CI 1.06-3.33]); vintage >6 months (OR 2.43 [95% CI 1.42-4.16]), body mass index >35 kg/m2 (OR 1.94 [95% CI 1.18-3.20]), post-dialysis systolic blood pressure (SBP) <115 mmHg (OR 1.57 [95% CI 1.04-2.37]) and intradialytic cramps (OR 1.76 [95% CI 1.09-2.86]). There were no associations with gender, race, age, ESRD etiology, intradialytic SBP <90 mmHg, serum sodium or potassium, dialysate sodium, bicarbonate or temperature, blood flow rate, or ultrafiltration rate. CONCLUSIONS: Responses to the recovery question were reliable and low dialysate potassium was associated with prolonged recovery.