PREDICTION OF EQUILIBRATED POSTDIALYSIS BUN BY AN ARTIFICIAL NEURAL-NETWORK IN HIGH-EFFICIENCY HEMODIALYSIS

In urea kinetic modeling, postdialysis blood urea nitrogen (BUN) is us ually underestimated with an overestimation of the Kt/V especially in high-efficiency hemodialysis (HD). Thus, an artificial neural network (ANN) was used to predict the equilibrated BUN (C-eq) and equilibrated Kt/V (eKt/V-60) by using both predialysis, postdialysis, and low-flow postdialysis BUN. The results were compared to a Smye formula to pred ict C-eq and a Daugirdas' formula (eKt/V-30) to predict eKt/V-60. Seve nty-four patients on high-efficiency or high-flux HD were recruited, T heir mean urea rebound was 28.6 +/- 2%. Patients were divided into a ' 'training'' set (n = 40) and a validation set (n = 34) for the ANN, Th eir status was exchanged later, and the two results were pooled, In th e prediction of C-eq, both Smye formula and low-flow ANN were equally highly accurate. In patients with a high urea rebound (>30%), although Smye formula lost its accuracy, low-flow ANN remained accurate, In th e prediction of eKt/V-60, both Daugirdas' formula and low-flow ANN wer e equally accurate, although the Smye formula was not so accurate, In patients with a high urea rebound, although both Smye and Daugirdas' f ormulas lost their accuracy, low-flow ANN remained accurate, We conclu ded that low-flow ANN can accurately predict both C-eq and eKt/V-60 re gardless of the degree of urea rebound. (C) 1998 by the National Kidne y Foundation, Inc.