Explaining counter-intuitive clinical outcomes predicted by Kt/V.

Abstract

Population-based studies of maintenance hemodialysis patients have demonstrated a reproducible relationship between the dose of hemodialysis and mortality and morbidity outcomes. In these analyses, which have aggregated hemodialysis patient subgroups, improved outcomes are associated with greater doses of hemodialysis. However, remarkable counterintuitive findings are observed if patients are analyzed by subgroups based on their race, gender, and anthropometric and blood-based biomarkers of nutritional state. For example, blacks generally receive lower doses of hemodialysis than whites, but enjoy relatively improved survival; patients who receive the highest doses of hemodialysis have an increased death risk; and the dose response curve between hemodialysis and survival is altered based on the patients' body mass index. These seemingly paradoxical relationships between hemodialysis dose and patient survival can be explained because of the use of mathematical urea kinetic constructs as clinical outcome predictors; they integrate a measure of solute removal (K x t) with an anthropometric surrogate of nutrition, the urea distribution volume (V). Both these measures have an independent influence on patient survival and in some clinical circumstances are of unequal power as clinical outcome predictors. These complex interactions must be kept in perspective as clinical care is delivered in the context of hemodialysis dose.