Dr. Carl Kjellstrand
Recipient of the Annual Award for Lifetime achievement in Hemodialysis


DE banner


keynote address




The artificial kidney should try to mimic normal renal function as much as possible. The main function of the kidneys is the even maintenance of the internal milieu, homeostasis, not only the removal of urea. We should strive to maintain normal electrolyte and acid-base balance and normal fluid volumes in the body. This is achieved with long, daily, smooth dialysis. The removal of urea is one, and not the most important, of many kidney functions, yet we dialyze as if the only renal function was urea removal.


Chronic hemodialysis in the USA has been a quantitative triumph and a qualitative failure. From a handful of people being dialyzed at the end of the 1960s, an enormous organization has been built up that accepts over 100,000 patients annually — 1 every 5 minutes — and cares for over a quarter of a million people. It is an impressive story, and one should be proud to have participated in it.


The success of chronic hemodialysis as measured in survival, morbidity, and quality of life, is another, sadder story. Data from the early European Dialysis Transplant Association (EDTA) Registry indicated that the 1-year crude mortality rate fell from about 50% in the 1960s to 10% in the early 1970s. The US National Dialysis Triangle Registry reported a cumulative 1-year mortality in 1970 of only 10%. Since that time mortality almost tripled to 25% in 1990, and has since shown a modest decline to about 23%. It is better to develop cancer of the breast or prostate than to need chronic dialysis in the USA in the year 2000! The increasing death rate is unprecedented. In all other diseases employing high technology medicine, such as coronary bypass, HIV, hypertension, and diabetes, mortality rates have improved. One explanation for the rising mortality in the USA has been that older patients with more comorbidity and worse diagnoses are accepted. The same is true for Japan, second to the USA in accepting patients for dialysis, but their mortality rate, which bottomed out at 9%, has only recently risen to 10%. All other high technology treatments are also accepting older and sicker patients, such as those with dia-betes; still, the results have improved. Dialysis is unique in experiencing a worsening mortality. I believe the major reason for the miserable results of chronic dialysis is that we dialyze in an increasingly unphysiologic fashion.


If one considers the reasons for these awful statistics in the USA, they can be divided into patient problems that limit survival and problems with the hemodialysis procedure that should be improved. Advanced age and many comorbid conditions limit survival, but when the patient comes for dialysis, we can do nothing about these. With hemodialysis there are four areas where improvement can be achieved: higher small molecular removal (higher Kt/V), more-physiologic dialysis, better middle molecule removal, and better biocompatibility. The best clinical studies indicate that the last two are of lesser importance, and that more-physiological dialysis is the best way to improve mortality. A higher Kt/V is the second most important factor.


Dialysis has become increasingly unphysiologic. The early hemodialyses were long affairs, 8 –12 hours. In the USA, because of commercial pressures, and in other countries because of budget constraints, dialysis was increasingly shortened to the current 2 1/2 – 3 1/2 hours, only one fourth as long. Because of the preoccupation with dose (Kt/V), it was erroneously assumed that as long as the result of the numerator, K x t, was maintained, all was well. This assumption is wrong. A greater K cannot compensate a short t. A long t is necessary for fluid and sodium removal, and thus, blood-pressure control. Short, ultrafast dialysis can never maintain the homeostasis necessary for extended survival and decreased morbidity. In Japan, as in the USA, short dialysis is dangerous. Overall, Japanese patients have only one third the risk of dying in a given time period compared to USA patients; however, the Japanese patients who dialyze as short and fast as USA patients, have the same awful death rate as in the USA. In the Tassin, France, clinic with the world’s best survival, patients are still dialyzed 8 hours. Practically none of the patients is on medications for high blood pressure, compared to some 70% of all USA hemodialysis patients. Homeostasis cannot be maintained with short, fast dialysis. Just as speed on the road kills, so it does in hemodialysis.


Short, fast dialysis also started a vicious cycle. Short, fast dialysis leads to more troubles than long, slow dialysis. Patients experience cramps, nausea, vomiting, headaches, and fatigue. Patients then naturally demand shorter and shorter dialysis so as not to have this misery prolonged. It will be necessary to explain to them that this short and fast dialysis is very dangerous to their survival, and longer, gentler dialysis will eliminate these symptoms. Patients also benefit from higher Kt/V dialysis than is now given. The Dialysis Outcomes Quality Initiative (DOQI) guidelines of a Kt/V of 1.2 – 1.3, three times per week, is far from optimal. Data from Park and Keshaviah and from the Japanese dialysis registry show continued survival improvement until a Kt/V of 1.6, and the healthiest dialysis patients are those on long nocturnal dialysis with weekly Kt/V of up to 12! These patients need neither blood pressure medications nor phosphate binders, and no or much less erythropoietin. Most of these expensive medications with unpleasant side effects are required due to infrequent and insufficient hemodialysis. Most of the complicated dialysis equipment with “sodium ramping or profiling” and hematocrit monitoring to avoid shock are necessary only because of dangerous fast dialysis. This is absurd!


It was clear in the first 2 years of chronic hemodialysis that increasing the frequency of dialysis was very important. Dialysis for vital indications and once or twice weekly dialysis were quickly abandoned for three-times weekly dialysis, but the logical conclusion of daily hemodialysis as the best dialysis was not drawn. The logistical pressures of a great increase in the number of patients, budgetary constraints, and profit motives were the reasons for this. However, since the early days of chronic dialysis, there have always been pioneers, trailblazers, and heroes of dialysis who have not been satisfied with the poor results of chronic hemodialysis as generally practiced. Since that time, close to 100 articles and abstracts dealing with daily hemodialysis, from over 30 clinics and including observations of over 250 patients, have uniformly showed improvement in hematology, blood pressure, nutrition, hormone levels, phosphate control, problems during and between dialysis, quality of life, hospitalizations, and rehabilitation.


The world’s best long-term survivals are those achieved in Tassin, France, with very long, slow dialysis, and with an average Kt/V of 1.79. A young patient in Tassin has only one fourth the risk of dying during the first 10 years of dialysis, compared to a young dialysis patient in the USA. The result for older dialysis patients, particularly vulnerable to the many side effects of short, hypereffective dialysis is even more astounding. A 65-year-old dialysis patient in the USA has a mortality risk 12 times greater than a 65-year-old patient in Tassin. The preliminary data from long, nightly dialysis, the most effective and physiological dialysis, show even better survival. The very preliminary survival results of shorter, daily hemodialysis appear to be equal to those in Tassin.


Considering the aggregate of all these data, it is clear that we are not giving our patients the best dialysis. Why should we be satisfied with three dialyses per week when each added dialysis leads to improvement? We do not treat hypertension every other day. We do not stop insulin on diabetic patients on the weekends; yet we dialyze, and therefore, treat the hypertension of dialysis patients only every other day, and we let patients go without dialysis for 3 days over the weekend, when many develop life threatening fluid overload and hyperkalemia. Dr. Teschan pointed out in the 1950s that one should treat these conditions in patients with acute renal failure before they became life threatening. The result of low efficiency CAPD, the most physiologic dialysis form, teaches the same. Frequent, long, smooth dialysis is the best dialysis, but our standard is infrequent, fast, and insufficient.

It is self evident that daily dialysis, and particularly long nightly dialysis, should be the standard rather than the exception. Absurdly, there are bureaucratic and payment obstacles in the way of giving the very best dialysis: payment for any dialysis sessions over 3-per week has to be justified. It should be the other way around. If we do not dialyze our patients daily, some explanation needs to be given for the substandard care provided them. While we should be proud of the fact that the USA has led the way in liberal and humane acceptance of dialysis patients — we have truly been the Statue of Liberty to them — we must now also strive to achieve the best survival.