Bedside formulas for K.t/V. A kinder, gentler approach to urea kinetic modeling

Abstract

Based on an empiric analysis of urea kinetic modeling equations, three formulas were derived for use at the bedside: (F1) Ob.t = 14.W.(K.t/V); (F2) target R = e-K.t/V + 0.03 + UF/W; (F3) K.t/V = -In (R - 0.03 - UF/W); where Qb = blood flow (ml/min); t = session length (h); R = post/pre BUN; UF = ultrafiltrate volume (L/session); and W = postdialysis weight (kg). When a particular K.t/V is desired, F1 is used to estimate Qb for a given W and t. Qb, t, and/or dialyzer type are then adjusted until R = target R calculated by F2 (based on target K.t/V, UF, and W). F3 estimates delivered K.t/V from R, UF, and W. These formulas were validated against 336 conventional 3-point modeling sessions in 256 patients. The multiplier "14" in F1 actually averaged 12.5 +/- 2.8, the large standard deviation suggesting that use of F1 would often result in K.t/V values substantially above or below the target K.t/V. The main causes of error were unusual V/W or K/Qb. On the other hand, F2 and F3 were highly accurate: target R estimated by F2 predicted actual R (at a given K.t/V, UF, and W) very well (% error = 0.65 +/- 2.3); K.t/V estimated by F3 predicted actual K.t/V (% error = 0.74 +/- 2.7). The results suggest that attainment of a target R (F2) will ensure that the target K.t/V is being delivered; F3 is useful to estimate K.t/V.F1 is an initial estimate only of the dialysis Rx; it must be followed by adjustment of t, Qb, or dialyzer until target R (from F2) is attained.