By Peter Laird, MD
The NCDS introduced urea kinetics to the dialysis world as the most important measurable factor in the prescription of dialysis even though there was no financial disclosure by the authors including the fact that Edmund G. Lowrie was then the Vice President of National Medical Care Inc., the largest for-profit dialysis corporation in America at that time. Since the publication of the secondary analysis in 1985 introducing the Kt/V concept by Gotch and Sargent, dialysis outcomes in America have plummeted to the lowest survival of any in the industrialized world. As Gary Peterson noted in today's lead story on RenalWeb, the discussion and controversy continues unabated as seen in a recent editorial in MedScape by Joel D. Glickman, a respected nephrologist who has served as a consultant for both Davita and NxStage.
It all started with the NCDS published in the New England Journal of Medicine in 1981. This study randomly assigned just 151 hemodialysis patients into 4 treatment groups to measure the effect of 2 variables on survival: time on dialysis (short or long) and time average urea concentration (high or low). The authors of the study concluded that morbidity was worse in the group with high blood urea nitrogen, but time on dialysis was not a significant determinant of outcomes. Patients who received longer dialysis treatments had a trend to less time to first hospitalization, but the P value was .06.
Commenting on the article, Chertow and colleagues said, "It is difficult to conclude from the NCDS that session length is not meaningful; the authors presciently stated that the study conclusions should be interpreted cautiously, given the design and limited powers of NCDS. In retrospect, one might argue that the NCDS' session length P = .06 was the most significant (important) "nonsignificant'" (statistically) effect in the history of dialysis research." Regardless of the interpretation of the study results, we must recognize that the long treatment time was 4:30 + 3 minutes and the short treatment time was 3:15 + 3 minutes. The effects of shorter treatment times, such as the 3 hours K.R. was prescribed, were not studied.
Where could we have gone wrong when these results which were both significant and randomized in a controlled study? Simply put, Kt/V and urea kinetics from the NCDS not only ignores the influence of TIME in dialysis outcomes in what I believe was a flawed analysis, it completely ignores the founding principles of optimal dialysis: the Middle Molecule Hypothesis by Dr. AL Babb, et al (here) and Dr. Carl Kjellstrand's Unphysiology Hypothesis (here), both known at that time. Urea is a small molecule easily cleared by dialysis that many argue has no clinical impact in the physiology of the uremic condition. It is a marker but probably not a direct player in the damage done by uremia. Better dialysis markers are problematic to measure such as B-2 microglobulin which has complicated kinetics that are TIME dependent. Fortunately, many researchers are revisiting the middle molecule theory identifying several of the toxic molecules of uremia:
Clinical dialysis studies should from now on be focused on more efficient, convective removal strategies, first searching for the most adequate applications and then evaluating their clinical impact. The next step then should be to develop more specific removal strategies (e.g., adsorption) aimed at the compounds which are classified as the most important ones. This might emanate in less costly removaloptions than the empiric strategies applied as of today, and such systems might also be regenerated, hence enabling an earlier start of removal than currently applied.
The measurement of urea and the subsequent mathematical mental masturbation we call the Kt/V that makes researchers, clinicians and patients feel good about the "adequate" dialysis they deliver fails to have an underlying theoretical physiologic basis other than the transfer of urea. High Kt/V's can be manufactured in dialysis units by manipulating ultra-filtration rates, blood flow rates and other factors that foreshorten dialysis treatments especially in patients with a small total surface area and total body volume such as many petite female patients who often receive the shortest dialysis sessions of any patients, yet have the highest Kt/V's paradoxically. Kt/V measures the clearance of a clinically unimportant molecule masquerading as the arbiter of dialysis outcomes.
Dr. Kjellstrand's Unphysiology theory should instead also be one of the guiding factors in improving dialysis outcomes through more frequent and longer duration dialysis thus eliminating the unphysiology of dialysis from infrequent, rapid and short dialysis sessions common in American dialysis units today. The NCDS neglected this theory as did the HEMO study two decades later bringing death by randomized and controlled study from adherence to the failed concepts of Kt/V. Each month, dialysis patients anxiously await the results of blood tests. Each month, the renal staff and physicians pronounce their dialysis adequate by a magical wand without dimensions called the Kt/V. Yet, each month, these same patients die at a rate 2.5 times that of patients undergoing chronic hemodialysis in Japan. This is the cost of ignoring Dr. Kjellstrand and Dr. Babb for over 35 years:
In the early 1970s, Kjellstrand et al. [12, 13] investigated factors related the untoward effects of dialysis. They showed that large fluctuations in body weight, electrolytes, osmolality, and urea concentration were more important with respect to morbidity than concentrations of small uremic toxins or middle molecules. This led them to formulate the ‘unphysiology’ hypothesis which suggested that wide swings of solutes and fluids in the body were significant causes of morbidity in dialysis patients. Daily or continuous dialysis would more closely mimic the function of the native kidney by reducing the magnitude of solute and fluid oscillations and so would be expected to be superior to the usual intermittent dialysis regimens.
The deception of "adequate" hemodialysis is at the center of our horrific American dialysis mortality rates now debated for the last four decades, yet so little is done to correct this situation. As much as I appreciate the commentary by Dr. Joel Glickman to focus on longer dialysis treatments, I believe that the error of his opinion is to continue to utilize the current urea kinetics model as the basic standard of care. Putting the 't' back in Kt/V will only impede implementation of physiologic based dialysis treatments. Nephrology is perhaps the most highly dependent specialty based on physiology. Ironically, dialysis, the most recognized nephrology procedure ignores the unphysiology of dialysis for the majority patients nor does it correct the middle molecule millieu of uremia.
I would hope nephrology will one day unite and recognize that Scribner, Kjellstrand and their colleagues were right all along. I suspect that pride shall prevent that epiphany for many years to come at the cost of hundreds of thousands of lives unless the patients themselves one day learn enough to refuse the substandard conventional American dialysis protocols. The popularity of medical blogs coupled with market pressures from new and innovative dialysis machines is more likely to become the instrument of change than decades of reason and logic. What a sad commentary from a discipline that pays frightful attention to the most minute changes of urine physiology thereby diagnosing complex renal disease from human waste, while at the same time ignoring the basic physiology of their most important medical procedure rendering their patients lives wasted and disposable to the system. Perhaps, instead it will be the patients who one day unite and overcome the enslavement of the large dialysis organizations that use our pain and suffering to their advantage. Perhaps one day.