Managing the Trajectory of Kidney Disease: Priorities in Dialysis Care

Eric Cohen, MD


August 17, 2005

Priorities of Care

The multiple issues related to the care of chronic dialysis patients make it difficult to quantify what's very important, what's less important, and what's not so important in their clinical management. The short-term imperatives of a failed vascular access demand a different approach and a different time frame from the consideration of the level of phosphorus or blood pressure (BP) control in a stable, long-term dialysis patient. Bearing in mind this variability, I attempt to identify the major key priorities in dialysis care, and differentiate short-term from long-term concerns. The focus will be on hemodialysis.

Before reviewing the priorities of care, it is worth considering that patients with chronic renal failure (CRF) who do not require dialysis have adjusted mortality rates that are inversely related to their level of kidney function. That is, mortality rates increase as the serum creatinine (SCr) level goes up, and for an estimated glomerular filtration rate of 15 mL/minute, this rate is about 15%.[1] With this in mind, the average annual mortality rate of 20% in American dialysis patients becomes more understandable (Figure 1). Given the existing high mortality rates of never-dialyzed patients with CRF and the implied underlying burden of disease that these patients carry with them as they start dialysis, the difficulty of improving long-term dialysis survival outcomes becomes apparent.

Death rate, per 100, chronic renal failure and dialysis patients.[1] The average annual mortality rate in American dialysis patients is 20%, which is just beyond the approximate 15% rate for individuals with severe, predialysis chronic renal failure.

Good dialysis delivery is a sine qua non of long-term survival for a hemodialysis patient, and that can only be achieved by proper access to the circulation. Long-term, arteriovenous (AV) fistulae are clearly superior to AV grafts, which are better than catheters. On average, a hemodialysis catheter will become infected within a year of insertion.[2] Proactive access management, including timely referral to access surgeons and venous mapping, appear to lead to higher rates of AV fistula placement.[3]

Once in place, AV fistulae require little maintenance. AV graft patency is not helped with warfarin, but secondary patency of AV grafts may be improved with aspirin.[4] Access monitoring should not require Doppler flow studies, because venous pressure monitoring is sufficient and cheaper.[5] At a blood flow rate of 200 mL/minute, an AV fistula or graft venous pressure > 140 mm Hg points to venous limb stenosis, and fistulography should be done. Prolonged dialysis bleeding may also point to a venous stenosis, and that may occur without elevation in venous pressure in the case of an AV fistula.

Good dialysis delivery is important in any form of dialysis, hemodialysis, or peritoneal dialysis -- acute or chronic. Measurement of dialysis delivery has been analyzed and debated for decades. Although the urea reduction ratio (URR, predialysis urea minus postdialysis urea/predialysis urea) has shortcomings as a marker of dialysis delivery, it is the accepted standard of the care of chronic hemodialysis patients. The URR is a measurement for a single dialysis treatment and is commonly monitored monthly (ie, 1 of 13 dialysis treatments). A URR of 0.65 is the minimum standard, and 0.7 is recommended as a safety margin.[6] These values are equivalent to 65% and 70%, respectively, when expressed as percentages.

Missed or shortened dialysis treatments, which reduce the overall dialysis adequacy, occur in more than 20% of all dialysis patients.[7] The clinical definition of dialysis adequacy is not well quantified, but uremic symptoms can be used as a guide. This is especially true of uremic itching, which is worse when dialysis delivery is decreased[8] and improves when dialysis delivery is increased.[9] In general, one should strive to eliminate all uremic symptoms, with the goals being no itching, no shortness of breath, and a good appetite.

Enhanced dialysis delivery may be clinically indicated, regardless of the URR. The question of how much of an increase in the URR or its correlate, the Kt/V (dialyzer clearance x dialysis time/body water volume), is beneficial in terms of better patient outcomes remains unanswered. The Hemodialysis (HEMO) trial suggested that a Kt/V > 1.3, which is equivalent to a URR near 0.7, was not beneficial in terms of improving patient mortality compared with a Kt/V of 1.3.[10] However, this trial was flawed because it studied a cohort that had been on dialysis for an average of 3 years at the beginning of the study and that had a baseline mortality of 16%. Neither characteristic is typical of the American chronic dialysis population. In a national study of more than 45,000 prevalent dialysis patients, Port and colleagues[11] showed that mortality decreased as the URR increased from 0.65 to 0.75. This study population consisted of unselected dialysis patients, which are representative of real-life experience, rather than the selected patients in the HEMO trial. In addition, a similar analysis showed that better Kt/V was associated with less cardiovascular and infectious mortality.[12] Such cohort studies, even if retrospective, are probably more useful for clinical practice than the HEMO study.

The current American dialysis population has a much higher median age (65 years) than the general population (35 years). In addition, the vintage distribution (time since initiation of dialysis) is not a bell curve, but rather a highly skewed distribution, being greatest for the least time on dialysis and tailing off rapidly for more time on dialysis (Figure 2). This can be explained by death of patients on dialysis, and to a much lesser degree, successful kidney transplantation. It thus stands to reason that later complications of chronic dialysis, such as acquired cysts of end-stage kidneys, are of much less significance than the ever-present concerns about access and dialysis adequacy.

Dialysis vintage. More than 70% of patients are on dialysis 3 years or less.[20]

Better dialysis delivery is associated with longer survival, so that long-term concerns in dialysis care may be less important for not-so-well-dialyzed patients and more important in well-dialyzed patients. This may explain the disparate interpretations of the role of hypertension in long-term dialysis survival. In a dialysis population with a Kt/V of about 1, Salem and Bower[13] found no benefit of normal BP. In a much better dialyzed population with an average Kt/V of 1.7 (UUR of .75), Charra and colleagues[14] found a distinct and impressive benefit of BP control, aiming for a predialysis mean arterial BP of 100 mm Hg or less. Further analysis by the latter group showed that the benefit of controlling the BP was only apparent after several years. Thus, such delayed benefit, which is similar to the delayed benefit of BP control in the general population, is only realized if dialysis patients survive long enough.

Elevated parathyroid hormone (PTH) blood levels are present in almost all dialysis patients. PTH and its determinants have received much attention in recent years, not only because of improved testing but also because of new treatments. Making sense of the data is difficult, however. Dietary phosphorus, dietary calcium, dialysate calcium, vitamin D analogs, phosphate binders, and the calcimimetic agent cinacalcet are the main players in this arena.

The relevance of secondary hyperparathyroidism (SHPT) is no longer limited to symptomatic bone disease, which is now uncommon. SHPT is now thought to possibly play a role in cardiovascular calcification.

It is worth noting that large cohort studies have shown that of PTH, calcium, and phosphorus levels, a high phosphorus level poses the greatest relative risk for mortality.[15] In addition, the benefit of controlling the PTH level per se has not been shown in terms of morbidity or mortality. In this report by Block and colleagues,[15] the increased risk of mortality related to hyperphosphatemia was shown in patients with a dialysis vintage of < 2 years, which is clearly a relevant group of patients.

It is also important to note that the hard end point of parathyroidectomy, an operation done for refractory SHPT, occurs at a median time of 6 years after the start of dialysis.[16] Less than 15% of dialysis patients survive that long. Stated another way, worrying about excess PTH levels and the possible future need for parathyroid surgery is a long-term concern in only a small fraction of patients. Thus, our clinical priority should be to control the phosphorus level, not the PTH level. Keeping the predialysis phosphorus level ≤ 5.5 mg/dL will just about guarantee that the PTH level will not be out of control. In this regard, phosphate is dialyzed, and good dialysis will help to control the phosphorus level.

Multiple studies have shown a clear and compelling association between anemia and increased morbidity and mortality in chronic dialysis patients. The Kidney Disease Outcomes Quality Initiative (K/DOQI) guidelines state that one should strive for a predialysis hemoglobin level of 11-12 g/dL. Parenteral erythropoietin will be needed to achieve this target in > 90% of dialysis patients. However, 2 remarkable intervention studies in chronic dialysis patients have failed to show the benefit of anemia correction to normal levels of hemoglobin and hematocrit.[17,18] It appears prudent then to adhere to the current K/DOQI guidelines and to avoid overtreatment of anemia. Erythropoietin management protocols, managed by dialysis nurses with physician oversight, can be very useful.

The dialysis population is quite different from the general population. Dialysis patients are much older and have a high annual mortality rate. The medical priorities of a 65-year-old end-stage renal disease (ESRD) patient on dialysis are markedly different from those of a typical 65-year-old in the general population. Stopping dialysis in the ESRD patient for 10 days or more is likely to be fatal.

Chertow and colleagues[19] have shown that screening for common cancers, such as breast or colon cancer, is a waste of time in dialysis patients. Not only is the gain in lifetime as a result of screening insignificant, but the money spent on screening would lead to much greater lifetime gains if it were spent on improving dialysis delivery.

It is not known whether dialysis patients benefit from having a primary care physician in addition to a nephrologist. The US Renal Data System 2004 Annual Data Report indicates that payments to internal medicine physicians, separate from nephrologists, represent about one fifth of the total payments to physicians in the ESRD program.[20] This may represent payment for "primary care."

As noted above, for common cancers, general medical screening is unhelpful in chronic dialysis patients. In addition, vaccinations are easily given at dialysis units, sparing the 3-times-a-week dialysis patient from the extra effort and expense of an additional visit to an internist or primary care physician. Thus, primary care for dialysis patients may actually be cost-ineffective. Nonetheless, it is possible that for an individual patient, a long-standing relationship with an internist or family medicine physician may provide a beneficial sense of continuity.

Finally, the day-to-day practicalities of a dialysis unit cannot be ignored, and are important for all patients, new or old. These include maintenance of water treatment and dialysis machines, and maintaining a well-trained nursing, ancillary, and technical staff.