Abstract Background and Aims Patient centred care is progressively gaining importance. Standardized outcomes should take into account patient relevant outcomes, such as mortality and quality of life. It is unknown whether the health utility index EQ5D, an objective assessment of quality of life, is associated with mortality in haemodialysis patients. Serum concentrations of different uremic toxins have been associated with survival. We intended to explore whether EQ5D rather than concentrations of representative uraemic toxins (UT) was associated with mortality. Method Prospective longitudinal multicentric cohort study of all haemodialysis patients at 5 representative dialysis centres in Flanders. Total and free concentrations of representative uraemic toxins indoxyl sulfate (IxS), p-cresyl sulfate (pCS), p-cresyl glucuronide (pCG), indole acetic acid (IAA), 3-carboxy-4-methyl-5-propyl-2-furanpropionic acid (CMPF), hippuric acid (HA) and uric acid (UA) were determined at baseline. Mini Nutritional Assessment (MNA) and EQ5D were assessed by dedicated nurses. Mortality was assessed from administrative or medical records. A Cox regression model was built, both fixed and as conditional forward including clinical parameters, MNA, EQ5D and serum concentrations of the different uraemic toxins. Results 216 prevalent patients (62%male, age 67.2 ± 15.7) all on high flux haemodialysis were included. Over the observation period, 59 patients (27%) died. Health utility as assessed by EQ5D was 0.69±0.28; perceived health by visual analogue scale (VAS) (0-100) was 59.6±18.6. Adjusting for age, gender and MNA score, free but not total concentrations of IxS, pCS and IAA and EQ5D score were associated with overall survival in the forced entry model. However, neither in the fixed entry, the forward or backward model were serum concentrations of uraemic toxins retained, and only age, gender, BMI, EQ5D and VAS were consistently associated with survival. Conclusion Measures of health utility, such as EQ5D, and quality of life, such as a simple VAS score, are more predictive of survival in patients on regular haemodialysis than serum concentrations of a panel of relevant uraemic toxins. Health utility and QoL are thus not only directly relevant to patients, but also indirectly through an association with improved survival. The concept of dialysis adequacy has been challenged over the last decade, mainly because it was focusing on small solute clearance. Different authors reported that concentrations of middle molecular and protein bound solutes were more representative. In our study, free but not total solute concentrations were associated with survival. However, adjusting additionally for health utility and/or health perception and/or assessment of nutritional status abolishes this association. Studies assessing interventions to enhance solute removal should not only include solute concentrations as outcome, but also consider EQ5D, VAS and nutritional status as important outcomes to enhance patient centeredness of their findings.
Read full abstract