PubMed Searches

A History of Hemodialysis Adequacy

Maintaining life, but not health

When hemodialysis treatments were first used to replace renal function, no one knew how much dialysis therapy was needed to keep the patients healthy. Doctors knew that they had to maintain the blood levels of certain substances, such as potassium, within a specific range to keep their patients alive. The early pioneers of dialysis quickly mastered the technology to effectively manage these essential tasks. However, after the patients had been on dialysis for a few months, doctors found it was a much bigger challenge to keep these early hemodialysis patients healthy in the long run.

Many of these early hemodialysis patients had frequent and devastating complications, such as severe infections and inflammation of the heart cavity (pericarditis). Doctors discovered they could reduce or eliminate these complications by dialyzing the patients longer.  Somewhat reluctantly, patients accepted these longer dialysis times as necessary for maintaining their long-term health.  Nephrologists, trying to shorten the 8-10 hour dialysis sessions and also to prevent patients from being under-dialyzed, began looking for an easy-to-measure value (or "marker") to help them determine when a patient was receiving an adequate amount of dialysis therapy.

Urea (or BUN) levels

Urea levels in the blood had long been used to assess kidney function. Patients were typically described as "uremic" when they had renal disease. Urea is a waste product that results from the protein that we eat, digest, and catabolize and is normally excreted in the urine. When doctors saw new patients with renal disease, they observed that the higher the patient's BUN level was, the more ill they were.

In those early days, the general consensus was that high levels of urea had a toxic effect. Therefore, many doctors thought that keeping the BUN levels as low as possible would be best for the patients. They knew that the more protein the patients ate, the higher their BUN levels were. Accordingly, many of the early dialysis patients were advised to strictly limit the amount of protein in their diets.

When should we measure BUN?

Urea is efficiently removed from the blood by dialysis. It is a small molecule that passes easily through the dialyzer membrane. It is also constantly produced by the body. As you can see, BUN values vary widely over the week. For a Mon/Wed/Fri hemodialysis patient, it's highest on Monday pre-dialysis (~90). It's lowest on Friday post-dialysis.

Shooting for BUN Targets

In the 1970s and early 1980s, a common practice was to prescribe hemodialysis therapy in order to attain a target BUN. Some doctors wanted the pre-treatment BUN never to exceed 80 mg/dl. Some thought that the urea level should be averaged over the course of the week and they used a Time Averaged Concentration (TAC) BUN value as a target. To try and hit these target BUNs, they adjusted the amount of time on dialysis, the blood flow rates, changed dialyzers, and issued restrictions on dietary protein. Using target BUNs seemed like a logical approach to prescribing hemodialysis. However, there were many patients who hitting these BUN targets and were still not doing well, and some displayed symptoms of being underdialyzed.

Questions No One Could Answer

It also became apparent that some patients who ate a lot of protein were healthier than other patients who strictly limited their protein intake. Why was this?  Why did patients who weighed the same and ate the same amount of protein require different amounts of dialysis therapy to stay healthy? Why were some patients who had pre-treatment BUNs of 100 perfectly healthy, yet others who had pre-treatment BUNs of 60 unhealthy and in need of more dialysis?

Doctors began searching for a better "marker" in the blood that could be used to predict patient outcomes. Since urea is a small molecule, some began looking at larger "middle molecules", thinking that these would offer a better indication of an overall "toxic effect" on the body. Some clinics tried dialyzing all the patients the same amount of time, such as four hours or more. Some doctors experimented with calculations for prescribing hemodialysis therapy that included the patient's weight and the dialyzer surface area. Unfortunately, in using all these methods, there were some patients who were hitting cautiously-set targets and were still displaying symptoms of being underdialyzed.

National Cooperative Dialysis Study (NCDS)

In the 1970's, the NCDS was funded to try to determine which dialysis therapies provided the best patient outcomes. Data from hundreds of patients was collected and this provided a huge database of information about dialysis patients for the first time. The NCDS study compared the different methods of measuring dialysis adequacy in use at that time, but it could not provide a definitive answer as to what was best.  However, the accumulated data did produce a later breakthrough. Two researchers, Dr. Frank Gotch and John Sargent (PhD), analyzed the study's database trying to find new common factors for those patients that were doing well (and for those patients that were doing poorly).

Using Urea Clearances

Using the data from the National Cooperative Dialysis Study, Gotch and Sargent separated the patients into two groups based on their symptoms:  those that were well and appeared adequately dialyzed and those that had complications and appeared underdialyzed. They found the data didn't make much sense until they invented a new way of measuring dialysis therapy. Their new method still utilized urea, but it didn't use a target BUN. Instead, it measured the volume of blood that was cleared of urea during a treatment and compared it to the amount of water in the patient's body. The end result was that Gotch and Sargent arrived at a simple, elegant formula for measuring dialysis therapy:

Kt / V        (pronounced: Kay Tee over Vee)

(The "K" stands for "clearance" of urea in milliliters per minute, "t" for "time" in minutes, and "V" for body water "volume" in liters. The calculation uses each patient's individual body water volume, so the value is "normalized". Patients who weigh the same amount can have vastly different body water volumes.)

The Questions are Finally Answered

Why were some patients who had urea levels of 100 perfectly healthy, yet others who had levels of 60 unhealthy and in need of more dialysis? Why did two patients who weighed the same amount need different lengths of dialysis treatments to stay healthy? The formula Kt/V effectively answered these questions for the first time. When Gotch and Sargent applied the Kt/V formula to the data they had for these patients, the healthy and unhealthy patients fell into two distinct numerical groupings. If the patient had a Kt/V value that was 1.0 or higher, they were doing well in terms of being adequately dialyzed. If they had a Kt/V value less than 0.8, they were underdialyzed and were doing poorly.

Urea Kinetic Modeling

This new approach became known as urea kinetic modeling. It uses Kt/V values to prescribe and measure dialysis therapy. It uses the results of two blood tests, pre and post treatment BUNs, in its calculations. Urea kinetic modeling includes protein metabolism analyses and provides care givers, as part of the Kt/V calculations, with the amount of protein that the patients are actually eating (It calculates the PCR - protein catabolic rate. Diet diaries are notoriously inaccurate). Another benefit of Gotch and Sargent's analyses was that it provided strong scientific evidence that dialysis patients were better off eating more protein, not less. As more data accumulated, it became apparent that reducing protein in the diet to keep the urea levels low was actually resulting in patients not getting enough protein to stay healthy (low albumin levels). Over the years, it also became apparent that there were additional long-term benefits for the patients in increasing their Kt/V values to 1.2 and higher.

Urea kinetic modeling requires inputting several treatment and patient variables and the use of complex mathematics. Due to this complexity, a programmable scientific calculator or a computer is required to perform the calculations. Around 1990, researchers were able to show a high degree of correlation between Kt/V values and urea reduction ratios (URR).  A URR can be calculated with simple algebra and only uses the same two blood tests as the Kt/V equations. While a URR is not as accurate as a Kt/V value, nor does it provide any information about the patient's protein intake, a URR value does provide an easy-to-calculate marker for dialysis adequacy. As an example, a Kt/V of 1.2 is roughly equivalent to a URR of about 63 percent. Like Kt/V, the higher the URR value, the better.

Here is a comparison of URR and Kt/V from the NIDDK web site.

A Word of Warning: Current medical literature suggests that persons with renal disease who still have some remaining renal function should limit their protein intake to preserve that remaining renal function. Dialysis patients generally have no remaining renal function. Please consult with your nephrologist and renal dietitian about the protein intake level that is most beneficial for you and your current condition.

Gary Peterson, February 2000

(Author's note: This quasi-history lesson is intended to introduce new dialysis personnel to the concepts of urea kinetic modeling and adequacy in a somewhat entertaining fashion (a la James Burke). This short summary is not comprehensive nor does it do justice to all of the work that was done on this topic over the years by so many individuals. Those interested in more thorough and complete accounts should consult the appropriate scientific journals.)