Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The Dialysis Prescription
Perhaps the most fundamental prescriptive question is how much dialysis is enough? On the surface, this question seems simple. However, this veneer of simplicity is misleading and belies myriad complexities and nuances. The question’s apparent simplicity stems from its bounded nature: on the low end, zero dialysis is insufficient and leads to death; on the high end, there exists a finite number of hours per week during which a patient can dialyze; the optimal amount of dialysis must lie somewhere between these two. However, in pinpointing the location of this “optimum,” one must confront a series of difficult questions. (1) What are the direct objectives of dialysis? (2) How does one incorporate patient preferences? (3) Is there a threshold of “enough,” or are relationships continuous? (4) How are these answers impacted by temporal changes in sentiment and available technologies?
Question 1 is philosophical in nature. It is unclear if a consensus answer exists among stakeholders. In fact, the question itself may be too is simplistic in that it does not account for the need to potentially tradeoff among objectives (e.g., maximizing survival versus creating the least burden on quality of life). Question 2 extends uncertainty further. If there be an answer to the first question, that answer exists at the level of the “population” and does not imply conformity or like-mindedness among all patients. Both questions 1 and 2 assume a threshold effect: that there is an amount of dialysis below which outcomes are uniformly bad and above which outcomes are uniformly good. However, question 3 reminds us that the effect of dialysis on the body need not observe a threshold effect, and the relationship may be more incremental. Ergo, the right question may not be “how much is enough?” but rather “when is enough enough?.” Finally, question 4 reminds us that whatever the correct answer is—assuming there is a correct answer at a particular time in a particular context—may evolve and change over time.
Perhaps because these issues are so complex and interdependent, research and practice in dialysis have found a need to invoke simplifying paradigms and ask more directive, research-ready questions with empiric answers. In this chapter, we will explore how this line of reasoning has evolved, what it implies for contemporary practice, and how it may change in the future. However, be reminded that even this—the best available line of empiric data and data-driven practice—is by nature a shadow on the wall of Plato’s cave, and the reader is encouraged to harken back to earlier, more fundamental questions which may not have objective answers but nonetheless point to truth.
Philosophy of Dialysis Adequacy
Science is quantitative, and clinical medicine, at least semiquantitative. Therefore, in determining how much dialysis is enough, one must define the means by which dialysis can be quantified. The National Cooperative Dialysis Study was the first rigorous attempt to do so. The trial was begun in the mid-1970s at a time where dialysis had recently become “universally” available in the United States and where recent technological advances had made possible more efficient delivery of dialysis. At this time, there was no systematic tracking of patient survival (the abysmal survival rates among dialysis patients would not be fully recognized until the late 1980s), but providers recognized that many patients continued to feel uremic despite treatment adherence. In this setting, the obvious question was: how does one provide dialysis to patients in order that they not be uremic—the latter operationally defined based on terminal withdrawal from dialysis or hospitalization events? Investigators considered two metrics of dialysis dose: time-average blood urea concentration (50 vs 100 mg/dL) and time on dialysis (2.5–3.5 vs 4.5–5 hours) and randomized patients accordingly using a 2 × 2 factorial design. The former metric was meant as an index of small molecular weight toxemia and the latter as a proxy for a host of other factors. Upon trial termination, time-average blood urea was clearly associated with patient outcomes, whereas time on dialysis did not bear statistical significance. With the benefit of hindsight, one may question whether a p value of 0.056 in the context of a trial that was stopped early and never met target enrollment was truly indicative of “no association.” Nevertheless, it was clear that lower time-average blood urea nitrogen (BUN) led to preferable outcomes. In addition, a urea-based metric benefitted from other obvious advantages: (1) it corresponded to a decade-long legacy of basic uremia research that had used urea as a marker, (2) it was readily measurable, and (3) a single unified measure of dialysis adequacy was more readily implementable than one that considered multiple domains (i.e., urea and treatment time). On this basis, small molecular weight solutes in general and urea, in particular, became the currency by which dialysis adequacy came to be measured, a legacy that persists through the present day.
Kt/V
Leveraging data from the National Cooperative Dialysis Study, Gotch and Sargent introduced the concept of Kt/V in 1985. Kt/V is the mathematical relationship between the rate of urea removal (K) times treatment duration (t) divided by the volume of distribution for urea (V). Kt/V was, in essence, part of the emerging urea-centric tradition but broke from the existing paradigm in two key respects. The first distinction was that Kt/V attempted to balance urea removal (the numerator, K*t) to physiologic need (the denominator, V). The second distinction is subtle but nonetheless important: it changed the paradigm by which urea was considered from how much is there in the body? to how much is removed during dialysis? In typical mid-1980s circumstances, this distinction would have been academic: technologies were reasonably homogenous, and nearly all hemodialysis patients were dialyzed according to a thrice-weekly hemodialysis schedule. However, as we shall see later, interim changes in dialyzer efficiency and greater use of quotidian hemodialysis regimens have necessitated subsequent adaptations.
To be technically precise, Gotch and Sargent’s original concept of Kt/V has subsequently been dubbed single-pool Kt/V (spKt/V). This name derives from the conceptual kinetic model through which the equations were derived, which are largely analogous to first-order elimination kinetics. Within the original paper, and more so subsequently in later papers, there are a number of mathematic equations by which spKt/V can be calculated. Perhaps the most popular of these is the Daugirdas formula, which states:
where R is the ratio of postdialysis to predialysis BUN.
The use of closed-form equations to calculate Kt/V has been largely supplanted by iterative computer algorithms—the most popular of which is termed “urea kinetic modeling”—the Daugirdas equation is nonetheless revealing as to the constituent components of spKt/V. The first term in the equation (quantitatively the most important) describes the amount of urea that is removed from the start to the end of a dialysis treatment; this is an alternative mathematical formulation of the urea reduction ratio (URR), which will be discussed in the next section.
The second term in the equation corrects for urea generation during the dialysis treatment itself. To understand this term, consider that the typical thrice-weekly in-center hemodialysis patient spends approximately 7% of the time (12 of 168 hours) on dialysis; therefore, one would anticipate that at least 7% of urea is generated during dialysis itself (perhaps more if one considers the catabolic nature of hemodialysis). The effect of this is that the change in body urea content will underestimate the amount of urea being removed, an effect for which this term attempts to compensate.
The third term in the equation accounts for urea that is removed convectively through ultrafiltration. Because urea is freely permeable across the dialysis membrane, it exists in essentially equal concentration in the blood as in ultrafiltrate. Therefore, although urea is removed in the process of ultrafiltration, this is not otherwise reflected in the blood urea concentration.
The latter two terms are necessary because the prevailing paradigm is to measure urea removal on the patient side (i.e., by changes in blood chemistry) rather than on the drain side (i.e., by analysis of spent dialysate). (This is the only practicable solution: it is far easier to collect and analyze 14 mL of blood than 100 + L of spent dialysate.) However, as a thought experiment, consider a counterfactual circumstance where urea removal was assessed drain side: in this circumstance, no correction would be needed for urea generation or convective clearance. Both terms are similar in that they upwardly adjust spKt/V for factors not reflected directly in blood urea concentration change.
Urea Reduction Ratio
URR is a measure of the proportionate reduction in BUN over the course of dialysis. It is calculated as:
It is an alternative expression of the R term in the Daugirdas equation for spKt/V. For example, if a patient started dialysis with a BUN of 100 mg/dL and finished dialysis with a BUN of 30 mg/dL, R would equal 0.3, and URR would equal 70%. As thereby implied, there is an inherent mathematic correlation between URR and spKt/V. Discrepancies between URR and spKt/V derive from three primary sources: (1) URR is not directly indexed to body size, (2) URR does not directly account for urea generation during dialysis, and (3) URR does not account directly for convective urea losses. Nonetheless, the two are largely considered interchangeable, and both are endorsed by guideline committees as valid markers of dialysis adequacy.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here