Bhaskaran S, Tobe S, Saiphoo C, Manuel A
Calculating Kt/V using intradialytic urea values taken 30 minutes before the end of dialysis
43rd Annual ASAIO Conference, Atlanta
ASAIO J (Apr) 43:75 1997

One of the predictions of the regional blood flow model is, that post- dialysis urea rebound can be predicted by a simple formula:
Kt/V-dp = Kt/V-sp - 0.60 x (Kt/V-sp)/t + 0.03
James Tattersall, based on empiric observations from the standard 2-pool model, independently came up with another relationship: namely, that Kt/V-dp = Kt/V-sp measured exactly 35 minutes prior to the end of dialysis. Actually, the two are algebraically closely linked. For example, if the Kt/V-sp is 1.2 over 4 hours, then K/V is 0.3/hr, or 0.15 per 30 min, and rebound, or delta Kt/V will be 0.15 as predicted by the Tattersall method. With the rate equation, delta Kt/V is 0.6 x 0.3 - 0.03, or 0.15. So both equations will give an equilibrated Kt/V of 1.2 minus 0.15, or 1.05. Thus, it is not surprising that both the Tattersall method and the Daugirdas/Schneditz rate equation give very similar results. However, the rate equation should theoretically have an increased error, as its prediction of equilibrated post-BUN depends on both the post-BUN and the pre-BUN, whereas the Tattersall method relies on the post-minus-30 min BUN only.

In this paper, Bhaskaran et al measured post+30 min BUN (which they took to be an equilibrated post concentration) and compared estimates of this as predicted by the Tattersall method (using a post-minus-30 min BUN sample) and by the rate equation. They found that the "equilibrated" post-BUN was better predicted by the Tattersall method (r = 0.93) than by the rate equation (r = 0.79). The abstract is difficult to analyze, as key data are not given.

Comment: The data are consistent with statistical theory; i.e., that it is easier to predict a BUN concentration 30 min after dialysis with a single-BUN 30 or 35 min prior to the end of dialysis than with a formula based on both the pre and post BUN, as with the latter, errors in two measurements are involved. Having said this, regression coefficients are a poor way to compare results, and a mean percent error or absolute value percent error would give a better comparison. Also, the n of cases is not known. One problem is, that with either method, great care must be taken in how either the post-30 min sample (Tattersall method) or post sample (rate equation) are drawn. A slow flow method must be used to eliminate access recirculation; otherwise, either of the methods will give a falsely elevated estimate of equilibrated post BUN. It may be more practical, however, in a busy dialysis unit, to draw the post-sample at the end of dialysis, than to draw a sample at precisely 30 or 35 minutes prior to the end of dialysis, and the Tattersall technique makes it fairly easy to "cheat"; e.g., delay the time at which the post specimen is drawn. Both methods are fairly accurate, and the good results with the blood sided methods suggest that alternative dialysate quantification methods to estimate equilibrated Kt/V may not be as useful as previously thought. (John T. Daugirdas, M.D., University of Illinois at Chicago)

To go back use the BACK button on your browser.
Otherwise click on the desired link to this article below:
43rd Annual ASAIO Conference, Atlanta
Basic hemodialysis : Adequacy, prescription, urea kinetics