HDCN Article Review/Hyperlink

Held PJ, Port Fk, Wolfe RA, Stannard DC, et al

The dose of hemodialysis and mortality

Kidney Int (Aug) 50:550-551 1996

Owen et al (N Engl J Med 320:1001-1006, 1993), in a retrospective cross-sectional study of about 15,000 NMC dialysis patients, showed that mortality was linked to the urea reduction ratio. The mortality risk decreased exponentially as the URR was increased, leveling off at a URR of about 65%, corresponding to a Kt/V single pool value of about 1.2. The question as to whether further increases in URR are beneficial has led to the NIH-funded HEMO trial, where a URR of about 65% is being compared to a URR of about 75% in a randomized, prospective trial.

In this paper from the United States Renal Data Systems, the effects of URR and Kt/V are assessed in a randomly chosen sample of US dialysis patients. Although the sample size here is much smaller than that of the Owen/Lowrie study, the strenghts of the analysis are the random selection of study population, and the control for a panoply of comorbid factors. The analysis was limited to patients undergoing 3x/week dialysis using bicarbonate dialysate.

The data are quite similar to the Owen/Lowrie NEJM paper. The mortality risk appears to decrease until a URR of about 65-70% or a Kt/V of 1.2-1.3 is attained, and then appears to flatten, although there is difficulty in assessing from this data whether or not further increases in the URR are beneficial. The Kt/V and URR are tightly linked, the only real difference between them being the extra Kt/V afforded by ultrafiltration. Hence it was not surprising that either measure was a good measure of adequacy. Interestingly, when dose of dialysis was controlled for, treatment time was not a significant predictor of mortality (although treatment times longer than 4 hours are rarely given in the US, and this finding does not necessarily conflict with the observation of excellent survivals in places like Japan or Tassin, France, where dialysis times longer than 4 hours are common or even the rule).

Comment: The data appear to confirm the wisdom of the NIH Consensus Development Conference, which suggests that the target URR for dialysis patients should be at least 65% or the Kt/V at least 1.2. Interestingly, in this data set, the concept advanced by Alan Collins, that diabetics may need even higher doses of dialysis, was not confirmed. Cross-sectional studies such as these (and this applies to the USRDS, the Collins study, and the Owen/Lowrie analysis) are fraught with error, as one never knows whether the effect one is measuring is due to the variable in question, or to some associated, but uncontrolled for factor. For example, it may be that in the Collins paper, diabetics with less macrovascular disease were able to achieve higher levels of Kt/V, and diabetics with less macrovascular disease may intrinsically have a higher survival rate. So in conclusion, the dialysis community awaits with great interest the results of the NIH Hemo trial, which should definitively tell us if even higher URR or Kt/V levels than those presently recommended by the NIH Consensus Development Conference are beneficial. (John T. Daugirdas, M.D., University of Illinois at Chicago)