Body size, dialysis dose and death risk relationships among hemodialysis patients

The normalized treatment ratio, Kt/V derived from urea kinetic models (UKM), is a commonly used measure of dialysis dose. This measure assumes that smaller patients with low volume of urea distribution (V) require proportionately less total treatment (Kt) than larger patients. The conclusion has been questioned because the UKM use assumptions that could make them invalid for accurately predicting a clinical outcome like survival. It is possible that a relationship exists between Kt and body size whereby a different Kt is required for different sizes. This study therefore explored the relationships among body size, Kt, and death risk focusing on possible interactions between Kt and size.

The sample included 43,334 patients treated on January 1, 1999. Survival time was modeled using Kt or body size groups to evaluate the shape of the risk profiles. Kt and the size measures were then evaluated together as continuous functions both in main effects (that is, Kt and size) and interaction models to see if the association of Kt with risk might be different for different sizes. The size measures were body weight, weight adjusted statistically for height, body surface area (BSA), weight divided by height (wt/ht) and the body mass index (BMI).

The log of risk decreased in rough linear fashion for Kt, weight, weight for height, and BSA. The log-risk relationships were "reverse J-shaped" for wt/ht and BMI. The main effects models suggested improved survival with increasing Kt and all of the size measures. Adding an interaction term increased the benefit associated with increasing Kt and for weight, weight for height and BSA at low values of Kt and size. A significant, positive interaction term mitigated those effects at higher values. Thus, the death risk penalties associated with reducing Kt among small patients were as great as or greater than they were among large patients. A similar pattern was observed for V. Adding the interaction to the BMI model destroyed the main effects, so that there was no significant association between risk and either Kt or BMI. A cross-categorical model of BMI and Kt, however, revealed improving survival with increasing Kt among both low and high BMI patients throughout the range of Kt.

Evidence supporting the intuition that smaller patients require proportionately lower dialysis dose than larger patients was not found. To the contrary, smaller patients suffer as much risk as or more risk than larger patients from reducing Kt. Deciding dialysis treatment using a Kt/V based intuition may lead to avoidable under-dialysis particularly among small patients.