KDOQI Update 2000

I. Measurement of Hemodialysis Adequacy


Regular Measurement of the Delivered Dose of Hemodialysis (Evidence)

The dialysis care team should routinely measure and monitor the delivered dose of hemodialysis.

Rationale Numerous outcome studies have demonstrated a correlation between the delivered dose of hemodialysis and patient mortality and morbidity.19,33,34,36,38-42,57 The evidence demonstrates that mortality among ESRD patients is lower when sufficient hemodialysis treatments are provided. Because there is poor correlation between the dialysis care team’s clinical assessment of hemodialysis adequacy and patients’ clinical outcomes, unnecessary risk is placed on the patient unless rigorous methods of evaluation are used. Clinical signs and symptoms alone are not reliable indicators of hemodialysis adequacy.115,116 To ensure that ESRD patients treated with chronic hemodialysis receive a sufficient amount of dialysis, the delivered dose should be measured and monitored routinely. Guideline 6: Frequency of Measurement of Hemodialysis Adequacy offers guidance to the dialysis care team about the appropriate frequency for measuring and monitoring the dose of hemodialysis for adult and pediatric patients.


Method of Measurement of Delivered Dose of Hemodialysis (Evidence)

The delivered dose of hemodialysis in adult and pediatric patients should be measured using formal urea kinetic modeling, employing the single-pool, variable volume model.


The HD Adequacy Work Group considered several issues regarding the definition of acceptable and preferred measures of the delivered dose of hemodialysis. These included:

1. The comparative accuracy of alternative methods;

2. The completeness of information provided by alternative methods (eg, does the method support calculation of normalized protein catabolic rate [NPCR], which provides an estimate of the dietary protein intake in steady-state; will the method account for the impact of residual kidney function on the delivered dose of hemodialysis and NPCR);

3. The availability of dialysis unit staff to properly collect blood samples and record information from the dialysis session, such as the type of dialyzer used, intradialytic weight loss, blood and dialysate flows, true dialysis time, etc; and

4. The time to record, enter, and process this information.

Urea is the substance that is most often monitored in clinical practice as a surrogate for measurement of the clearance of small solutes in general. Reasons for this are that urea is a small, readily dialyzed solute that is the bulk catabolite of dietary protein,47,83 constitutes 90% of waste nitrogen accumulated in body water between hemodialysis treatments,47,83 is easily measured in blood, and that the fractional clearance of urea in body water correlates with patient outcomes, such as mortality19,33,35,36,38,39,42,48,57 and morbidity.34,36,39 Conventional methods of quantifying the prescribed or delivered hemodialysis dose begin by estimating the difference in predialysis and postdialysis urea concentration by sampling a patients blood before and after a single dialysis session.

The dialysate collection method is an alternative approach for quantifying the delivered hemodialysis dose. In this approach, the total dialysate that passes through the dialyzer during a hemodialysis treatment is collected. The total mass of urea removed is then calculated as the product of the urea concentration and the volume of spent dialysate. This method has been considered by some investigators to be the gold standard for urea kinetic analysis.45,117-119 Advocates of this method emphasize the advantage of minimizing exposure of patients and staff to blood-borne pathogens. However, the HD Adequacy Work Group recognized that dialysate measurement techniques are not routinely available, are impractical to implement in most hemodialysis units, have not been examined in relation to patient outcomes, and may be associated with the exaggeration of systematic collection errors.58,120-123 For example, a 7% error in dialysate collection can result in a 20% error in the equilibrated Kt/V. Although, the HD Adequacy Work Group also recognizes that dialysate side urea kinetics are best characterized as an equilibrated model, the Work Group thought it was best to focus on single-pool models of urea removal. Therefore, the Work Group focused on blood-based measurements of urea removal.

To normalize for differences in the size and habitus of patients, a dose of hemodialysis (prescribed or delivered) is best described as the fractional clearance of urea as a function of its distribution volume (Kt/V). The fractional clearance is operationally defined as the product of dialyzer clearance (expressed as K and measured in liters per minute [L/min]) and the treatment time (expressed as t and measured in minutes); the volume of distribution of urea is expressed as V and measured in L. Kt/V may be determined by formal urea kinetic modeling (UKM) or by extrapolation from the fractional change in blood urea concentration during a dialysis session. The delivered dose of hemodialysis may also be assessed using the URR.

Based on a review of the literature published prior to and since the release of the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis, the Work Group agreed with the RPA’s conclusion that formal UKM, based on either two or three BUN samples, was the best method for routine measurement of the dose of hemodialysis for adult and pediatric patients. Of the single-pool, variable volume mathematical analyses for quantitation of urea removal during a single hemodialysis session, formal UKM was considered to be the most accurate and complete. However, it is the least simple to implement.

Recent literature suggests that only one alternative method for calculating Kt/V (Kt/V natural logarithm formula) and one other measurement of the delivered dose of hemodialysis (URR) should be considered for routine use in adults. Respectively, these are:

(1)Kt/V natural logarithm formula44

Kt/V = -Ln(R - 0.008 × t)+(4 - 3.5 ×R) ×UF/W (1)

in which Ln is the natural logarithm; R is the postdialysis BUN ¸ predialysis BUN; t is the dialysis session length in hours; UF is the ultrafiltration volume in liters; and W is the patient’s postdialysis weight in kilograms; and

(2) URR19

URR=100 ×(1 - Ct/C0) (2)

in which Ct is the postdialysis BUN and C0 is the predialysis BUN.

Formal urea kinetic modeling. Formal kinetic modeling provides a quantitative method for developing a treatment prescription for a specific patient. Because of the complexity of the formulae that provide the information for calculation of Kt/V by UKM, computational software is necessary to compute Kt/V using formal UKM. Formal UKM can be used to calculate the exact treatment time required to deliver a particular hemodialysis dose at specified blood and dialysate flows with a particular dialyzer. Formal UKM requires accurate measures of:

1. Predialysis and postdialysis BUN for the first dialysis treatment of the week and the predialysis BUN for the second dialysis session of the week in a thrice weekly hemodialysis schedule.

2. Predialysis and postdialysis weights at the time of the first hemodialysis treatment of the week.

3. The actual treatment time, ie, the exact number of minutes during which the hemodialysis treatment was delivered on the first dialysis treatment of the week (not the prescribed length of treatment time or the time elapsed between putting the patient on the machine and taking him or her off).

4. The effective clearance of the dialyzer as measured in the hemodialysis unit (not the in vitro clearance value reported by the manufacturer alone).

Two-sample hemodialysis UKM based on the predialysis and postdialysis midweek BUN has been described and validated for accuracy in comparison to classical three-sample UKM.120,124

Advantages of urea kinetic modeling. When rigorously performed, formal UKM is a reproducible, quantitative method that has several advantages for assessing the adequacy of hemodialysis.36,48,124

Formal UKM can be used to develop an individualized hemodialysis treatment prescription. Hence, using UKM, the dose of dialysis (Kt/V) can be based on patient-specific parameters, such as residual kidney function and NPCR.48,83,126 To develop a hemodialysis prescription, it is necessary to obtain a dialyzer’s urea clearance (K) for a variety of blood and dialysate flows in blood/water. In order to provide this information to the dialysis care team, formal UKM uses the dialyzers urea clearance (K) that is provided by the manufacturer (typically derived in aqueous solution in vitro). The computational software for formal UKM then extrapolates a KoA value for that dialyzer that can be applied with a range of blood and dialysate flows. It is important that the urea clearance be calculated in order for the blood/water clearance to more accurately reflect the volume from which urea is removed and into which urea is generated reflected by the NPCR.

Formal UKM calculates the volume of distribution of urea by a complicated mathematical iteration of two formulae that share common terms (see Appendix B). The first formula solves for the end dialysis volume, Vt. The other formula calculates the urea generation rate (G) between consecutive hemodialysis sessions. The computational software repeats these formulae until unique values for Vt and G are found that satisfy both equations. Urea distribution volume can also be derived anthropometrically (see Appendix C). Both the anthropometrically derived urea distribution volume (for normal body composition is approximately 0.58_ body weight) and the kinetic extrapolated volume of distribution for urea are related to the patient’s height, weight, gender, and age (age is a variable in the Watson and Chertow formulae only).127-129 This value becomes the initial V for the Kt calculation (clearance _ time).

Once these two parameters are fixed, the treatment time can be calculated easily as follows:

[(Desired Kt/V)(Urea Volume, L)/(K, L/min)] = Treatment time, min

Example: [(1.4)(35/.250)] = 196 minutes

Various treatment time and blood flow combinations may be calculated by formal UKM and considered by the hemodialysis care team to achieve a desired Kt/V. Thus, the use of formal urea kinetic modeling provides the hemodialysis care team with guidance about which specific parameters of the prescription to modify to achieve the target hemodialysis dose. Changes in the dialysis treatment time, choice of dialyzers, or operating conditions such as blood and dialysate flows can be calculated to meet the selected prescription goal. The impact of residual kidney function on urea clearance can also be considered (see Appendix D). The latter information is also important for rigorous calculation of NPCR.

Potential prescribing errors. The Work Group noted that prescribing a Kt/V using an assumed volume of urea distribution, or one derived from anthropometric calculations only, has limitations that may compromise patient outcomes by diminishing the accuracy of the Kt/V. Estimates of the volume of distribution of urea (V) that are based on surface area nomograms, or calculations, may underestimate V in very muscular individuals. The erroneous V will lead to a proportional error in the calculated required Kt, which in turn may lead to prescription of a dose of hemodialysis that is too low for these patients.

Errors in prescribing the dose of hemodialysis may also occur using an assumed dialyzer clearance. The use of inaccurate hemodialyzer urea clearance values (K), based on incorrect assumptions from the manufacturers about KoA,130 or unappreciated, deleterious effects of hemodialyzer reuse131 may lead to significant errors in the hemodialysis prescription. The assumed K may overestimate the true clearance of a given device due to lot-to-lot variability (see Guideline 11: Baseline Measurement of Total Cell Volume). Actual dialyzer clearance may be affected by inadequate monitoring of hemodialyzer reuse procedures (see Guideline 13: Minimum Required Total Cell Volume). The failure of the dialysis care team to independently calculate the delivery of the prescribed dose of hemodialysis may permit this deficiency to go unrecognized. The incorporation of an independent method for error checks in formal UKM allows for improved monitoring of the dialysis dose.

Finally, although the HD adequacy guidelines are based on the assumption that no significant residual kidney function exists, the HD Adequacy Work Group agreed with the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis statement that the patient’s residual kidney function should be accounted for in the hemodialysis prescription. A minority of patients with ESRD have significant residual kidney function (Kr) that, when unaccounted for, may result in an underestimation in the actual total delivered dose of hemodialysis and an underestimation of NPCR (see Appendix D). These effects are a consequence of unaccounted urea removal by endogenous kidney clearance, especially between hemodialysis sessions. Although it has been suggested that the contribution of residual kidney function to Kt/V can be computed by simply calculating the weekly kidney urea clearance and dividing by three, these assumptions are incorrect. Likewise, Kr cannot be simply added to Kt, as this will greatly underestimate the effect of Kr. The single-pool, variable volume model of urea kinetics requires that the combined effects of Kt and Kr be analyzed with both continuous (endogenous) clearance and intermittent clearance (hemodialyzer-dependent) occurring over the treatment cycle. This complicated mathematical task is best performed by the computational software that is necessary for formal UKM (see Appendix D).

Formal UKM provides a mechanism to check for errors in the delivered dose of hemodialysis. Formal UKM requires measurement of the BUN concentrations pre-hemodialysis and post-hemodialysis, the delivered treatment time, and urea clearance for those blood and dialysate flow rates. Most formal UKM computer programs assume all input data are correct; they calculate the unique volume of distribution for urea in the patient that fits those data. By comparison, the value of V derived from the patient’s anthropometric features is only an estimate of urea distribution volume based on body composition analysis.127,128 A discrepancy between the kinetically derived V and the V expected from anthropometric calculations is an indicator of potential technical problems with the hemodialysis session. For example, if the kinetically derived V is larger than the anthropometric V, it may indicate underdelivery of the hemodialysis prescription secondary to problems such as:

1. Blood flow from the angioaccess was low.

2. The performance of the dialyzer was inadequate.

3. Dialysate flows during the hemodialysis treatment were less than prescribed.

4. The dialysis machine was programmed incorrectly.

5. The hemodialysis treatment ended prematurely.

6. The predialysis BUN sample was drawn after initiation of hemodialysis.

In the case where the kinetically derived V is larger than the anthropometrically derived V, the delivered dose of hemodialysis will be less than prescribed dose. Possible causes of an error in the kinetically derived volume of distribution of urea can be surmised from the variables that are used in its computation. Less frequently, the kinetically derived V is smaller than the anthropometrically derived V, so that the delivered dose of hemodialysis is greater than the prescribed dose. Some situations in which the postdialysis BUN concentration is incorrectly low and the kinetically derived V is too small include:

1. The postdialysis BUN sample was drawn from the venous blood line.

2. The postdialysis BUN sample was drawn in the setting of significant fistula recirculation.

3. The postdialysis BUN sample was drawn following a very efficient hemodialysis in a patient with a small V (high K/V), resulting in urea disequilibrium in the body water compartment.

4. The postdialysis BUN sample was inadvertently diluted with saline.

Based on the erroneous assumption that the resultant large Kt/V is correct, the dialysis care team may inappropriately advise a reduction in some component of the hemodialysis prescription, such as the duration of the treatment or the blood flow. Furthermore, because such errors are extended to the calculation of the PCR, the NPCR will also be inaccurate. Inappropriate dietary counseling to decrease the dietary protein/caloric intake may result.

Therefore, the Work Group agrees with the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis that despite its computational complexity, calculation of the kinetically derived volume is useful. Because errors are unavoidable in assessing the hemodialyzer clearance, treatment time, or BUN concentrations, mathematically fixing the Kt and calculating the effective volume of distribution for urea allows for the evaluation of the additive effects of all these potential sources of error.

A discrepancy in the volume of distribution of urea should alert the dialysis care team to an unanticipated error in the delivered hemodialysis dose. Formal UKM provides an accurate mechanism to check for errors in the delivered dose of hemodialysis, and it provides the greatest support for continuous quality improvement (CQI) efforts in the delivery of hemodialysis. Optimal quality improvement efforts require that the processes of care affecting patient outcome be routinely measured, individual deficiencies be easily defined, and corrective actions be simply implemented. The rigor of formal UKM supports the prescription of an appropriate dose of hemodialysis and allows errors in the delivered dose of dialysis to be easily recognized. Furthermore, formal UKM facilitates the identification of the variable(s) in the delivery of the hemodialysis prescription that were problematic, if the delivered dose was not the prescribed dose. Therefore, the HD Adequacy Work Group recommends that formal UKM be implemented as a component of a complete CQI program for ESRD care.

Formal UKM permits calculation of the NPCR. The volume of distribution term may be used to calculate the NPCR as follows:

NPCR (g/kg/day) = (PCR, g/day)(mean V/0.58)

Use of the NPCR enables the dialysis care team to perform longitudinal analysis of the patient’s nutritional status.47,48 The NPCR can be used to identify patients who might benefit from counseling about their dietary protein intake. Furthermore, it can be used to determine whether the hemodialysis dose needs to be escalated because of sustained high protein intake. (Calculation and use of the NPCR are discussed in greater detail in the RPA’s 1993 Clinical Practice Guideline on Adequacy of Hemodialysis51 and in the NKF-KDOQI Clinical Practice Guidelines for Peritoneal Dialysis Adequacy.109)

Disadvantages of urea kinetic modeling. The disadvantages of formal UKM are logistical. For example, the complexity of the calculations requires the use of computational devices and software. Physical parameters, such as the K and V, are difficult to measure and to monitor, and the actual treatment time can be difficult to determine. In addition, the time required for the dialysis unit staff to accurately collect and adequately process all patient information to support these calculations can be significant in large dialysis units. Finally, although the cost of the computers and software is low, it is a factor for some dialysis centers. Despite these relatively minor limitations of formal UKM, the Work Group agreed with the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis about the necessity of formal UKM. It is the most rigorous method for prescribing dialysis treatment and evaluating the consistency with which the prescribed treatment is delivered to the patient.51

Measuring delivered dose of hemodialysis in pediatric patients. Formal 2- and 3-point urea kinetic modeling can be reliably performed in pediatric hemodialysis patients.132-137 Despite wide variability in body size, Kt/V calculations using formal urea kinetic modeling are accurate and reproducible in pediatric patients. To date, there have been no studies of the dose of hemodialysis as a predictor of outcomes in pediatric patients.136,138 However, there is no reason to assume that the diligence in monitoring or the achievement of benchmarks should be modified for children. Although a small recent analysis demonstrated that the measured dialysis dose was equivalent whether using formal urea kinetic modeling or the Kt/V Ln formula,136 the HD Adequacy Work Group favors use of formal urea kinetic modeling to derive the Kt/V.

Other methods for measuring delivered dose of dialysis for adults and children. Recognizing that computational software to support the calculation of the Kt/V using formal UKM may not be available to all hemodialysis providers, several alternate methods for calculating Kt/V were examined by the HD Adequacy Work Group. Of these, the second generation logarithmic estimate, described as the Kt/V natural logarithm formula, provides the closest approximation to the single-pool, variable volume Kt/V derived from formal UKM.44

· Kt/V natural logarithm (Ln) formula. The Kt/V Ln formula, which accounts for intradialytic volume changes secondary to ultrafiltration and the resultant convective solute transport (solute clearance via ultrafiltration), is accurate over the full range of single-pool, variable volume Kt/V values (range, 0.7 to 2.1).44,45,119,139 The Work Group endorses this method as the best alternative for dialysis providers who are unable or unwilling to perform formal UKM for their adult or pediatric patients.

However, the HD Adequacy Work Group does not recommend this method for primary use because it has several important limitations compared to formal urea kinetic modeling. First, the Kt/V Ln formula alone does not support the calculation of the NPCR. An NPCR can be derived from a nomogram that uses patient-specific parameters and the Kt/V Ln formula,46,120 or by a described equation.139,140 However, these approaches have not been subjected to extensive comparative analysis. Second, the Kt/V Ln formula does not permit the rigorous, quantitative analysis of the hemodialysis prescriptions described earlier for formal urea kinetic modeling. For example, if the delivered Kt/V is observed to be too low, the natural logarithm Kt/V does not provide insight into how therapy should be altered (eg, dialysis time, blood and dialysate flows, and dialyzer).

· URR. The HD Adequacy Work Group acknowledges the ease of calculation and resultant popularity of the URR.31,66 Of the three methods that the HD Adequacy Work Group considered appropriate for measuring the delivered dose of hemodialysis, the URR is the simplest to execute. The URR has been shown to be a statistically significant predictor of mortality for ESRD patients.19,40

However, the Work Group agreed with the limitations of the URR as a measure of hemodialysis adequacy as delineated in the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis. Most important among these is that the URR does not account for the contribution of ultrafiltration to the final delivered dose of dialysis, in contrast to formal UKM and the Kt/V Ln formulae.140-142 This is because the convective transfer of urea that occurs by ultrafiltration does not result in a decrease in the BUN concentration, although urea removal into the dialysate has occurred. The result is that the URR is less accurate in estimating the delivered dose of hemodialysis than the single-pool, variable volume Kt/V calculated by formal UKM. For example, a patient with a large ultrafiltration requirement (as a percentage of body weight) will have a Kt/V that may be 0.2 or more greater than a patient with the same URR, but no ultrafiltration requirement.142 The result is that, at any given prescribed URR, the actual delivered dose of hemodialysis may vary substantially because of different ultrafiltration volumes. As shown in Fig I-1, a URR of 65% may correspond to a single-pool Kt/V (Kt/Vsp) of as low as 1.1 in the absence of ultrafiltration or can be as great as a Kt/Vsp of ~1.35 when ultrafiltration of 10% of body weight occurs.

Fig I-1. Impact of solute clearance through ultrafiltration losses on the delivered dose of hemodialysis as measured by the Kt/V. The curves are derived from formal urea kinetic modeling assuming a 3-hour hemodialysis, no residual kidney urea clearance, and a volume of distribution of urea of 58% of body weight. Wt refers to net ultrafiltration losses as a fraction of final body weight. Reprinted with permission.142

Furthermore, errors in the delivered dose of hemodialysis may be particularly difficult to detect in the target range of URR of 65%52 where a curvilinear relationship exists between URR and Kt/V.142 As a consequence of this nonlinear relationship, modest alterations in the URR can result in large clinically significant changes in the Kt/V. For example, assuming ultrafiltration losses of 2% of body weight during hemodialysis, a decrease in URR from 70% to 65% corresponds to a reduction in the Kt/V from 1.32 to 1.15.

The likelihood of error in the delivered dose of hemodialysis as measured by the URR is increased by its inability to support calculations of V for comparison to anthropometric values. In addition, like the Kt/V Ln formula, correcting observed deficiencies in URR requires empirical modification of the components of the treatment prescription. Similarly, the URR does not support calculation of the NPCR and ignores the contribution of residual kidney function to urea clearance. Thus, while the URR is a practical measurement tool for epidemiological outcome studies,19,40 its relative inaccuracy, and the incomplete information provided, compromise its use as the sole measure of delivered doses of dialysis in individual ESRD patients. Dialysis doses measured by URR have not been validated for pediatric patients. Routine substition of this measure may thus be especially problematic for this patient subset. Children have a smaller body and relatively higher total body water and are therefore at increased risk for significant amounts of urea rebound.

· Kt/V derived from the percent reduction of urea (PRU). The HD Adequacy Work Group reviewed literature on the accuracy of Kt/V calculations from the PRU that were published since the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis. The PRU formulae were derived by linear correlation to the Kt/V derived from total dialysate collection and from formal UKM.117,118 However, because the Kt/V derived from the PRU can be incorrect by approximately 20% in comparison to more rigorous measures of Kt/V,44,143 the Work Group felt that these formulae were not adequate for routine clinical use in measuring the delivered dose of hemodialysis and concluded that they should not be used.

· Urea (or solute) removal index and alternative methods of measurement. A less common method of expression of hemodialysis adequacy, the solute (or urea) removal index (SRI), is defined as the percentage of total body urea nitrogen content that is removed by a dialysis treatment. This measure could not be evaluated by the HD Adequacy Work Group because of the lack of clinical experience with this method.116,119,123 There are other promising methods under study, such as the use of devices that directly quantify urea removal during a hemodialysis session as the measure of intradialytic clearance,122,144-146 but judgments about their accuracy and routine utility cannot be made at this time.

· Kt. Recent reports that have described counterintuitive relationships between hemodialysis dose measured as URR and/or uneqilibrated Kt/V and patients’ outcomes. First, when grouped by weight for height percentile, a progressive deterioration of survival was observed for patients with the greatest URR and the lowest percentile weight for height.148 Second, another report described increasing mortality risks when the URR exceeded 70%, and so described a reverse j-shape for the URR-death risk profile. A similar mortality profile has been depicted by other investigators.40,149,150,186,189,191 By adjusting for TBW measured by bioelectrical impedance, the reverse j-shape can be eliminated, so that mortality benefit is seen over the entire specturm of URR values.186 An explanation offered was that a low TBW was reflective of poor nutrition, which was not overcome by a greater dialysis dose. An alternative interpretation was that patients with low TBW experienced greater urea rebound and hence received a clinically deleterious lower dose of dialysis than was appreciated by URR measurments alone. Third, although blacks receive lower dialysis doses than whites, their adjusted survival on hemodialysis is better.6,31,55,57 Blacks with ESRD have greater serum creatinine and albumin concentrations and anthropometric attributes than whites.57,152 This disparity in survival is negated by multivariable models in which the serum creatinine concentration, a laboratory-based surrogate of somatic nutrition, is included.152 Last, using data from a US Renal Data System special study, the relationship between the equilibrated Kt/V, patients’ anthropometric attributes, and mortality was examined.154 Kt/V and body size were both independently and significantly related to mortality after adjustment for patients’ characteristics and comorbid conditions.

These findings have been interpreted as being consistent with an independent beneficial effect of nutrition on survival for hemodialysis patients, which is reflected in their anthropometric attributes and hence measures like the urea distribution volume. In the urea kinetic constructs for dialysis dose, such as Kt/V or URR, whereby one measure (K _ t) is divided by another (V, a proxy of lean body mass and nutrition), paradoxical relationships to outcome may be observed. For example, increasing the V, which is reflective of better nutritional health, may be associated with lower reported dialysis doses. This is unsurprising since many dialysis care teams prescribe relatively uniform dialysis doses with different patient weights. The findings reported above suggest that patient mortality is more reflective of the effect of V, than of the Kt/V.189 This hypothesis is supported by the finding that V, total body water, body weight, and body mass index are independent predictors of mortality risk for hemodialysis patients.189 Higher values predict better outcomes. Therefore, it has been suggested that disaggregating Kt/V and using the Kt alone offers a more clinically relevant strategy to measure the dose of hemodialysis.156,189 Based on cross-sectional data, minimum Kt thresholds of 40 to 45 L per treatment for female and 45 to 50 L per treatment for male ESRD patients have been suggested.156 Alternative theories have been posited to explain the more complex interactions described herein in which V is both an independent outcome predictor and the measure of the volume of distribution for the target solute, urea.158 The HD Adequacy Work Group does not feel there is sufficient evidence to advocate the substitution of Kt for Kt/V, but anticipate that the longitudinal design of the HEMO Study, with its well characterized patient cohort, will facilitate the examination of these complex interactions.92

· Duration of dialysis treatment. Some clinical researchers contend that the hemodialysis treatment time alone, independent of the Kt/V or URR, can be used as a measure of hemodialysis adequacy. Arguably, longer hemodialysis times may enhance patient survival.18,27,147 However, other investigators have observed that when the dose of hemodialysis was controlled, no relationship was demonstrable between the prescribed duration of hemodialysis and patient outcomes.19,34,40,149,151 Most of these studies have examined the prescribed treatment time instead of the delivered dialysis time; this discrepancy needs to be considered.153

The proponents of time as a measure of hemodialysis adequacy argue that the duration of hemodialysis is a surrogate for increased clearance of solutes other than urea that are not accounted for by urea-based kinetic models. Arguably, because this group of putative uremic toxins is much larger than urea,88,89,91,107 their clearance is less dependent on diffusion for blood-side removal, if conventional (low flux) dialyzers are used. Proponents also claim better control of other parameters that affect patient outcome, such as blood pressure.39,147,155,157 Conceivably, patients with longer hemodialysis treatments have fewer complications secondary to ultrafiltration and thus are more likely to routinely achieve their estimated dry weight. As a result, they suffer fewer cardiovascular complications secondary to hypertension and/or hypervolemia159-162 (see Guideline 15: Optimizing Patient Comfort and Adherence, and Guideline 16: Strategies to Minimize Hypotensive Symptoms). Some Work Group members felt that independent of the delivered dose of hemodialysis calculated using a single-pool, or equilibrated, variable volume model of urea kinetics, the duration of hemodialysis should not be permitted to fall below a particular threshold value (<2.5 hours). In the absence of empirical data, disagreement existed among the Work Group members with respect to this issue, and consensus could not be reached.

However, the Work Group agreed that the dialysis care team should be alert for the potential consequences of implementation of the prescribed treatment time derived from formal UKM, particularly for small-statured patients with a small V. If dialyzers with high urea clearance are used, the suggested treatment time to achieve the target Kt/V may be unacceptable to comfortably meet the patients needs for ultrafiltration (see Guideline 15: Optimizing Patient Comfort and Adherence). For example: (1) Patient V = 25L; (2) prescribed Kt/V=1.2; and (3) dialyzer urea clearance of 350 mL/min or 0.35 L/min, resulting in a treatment time of only 86 minutes (calculated based on 1.2 25/0.35). Such short treatment times in the setting of large K and small V may also result in inadequate delivered doses of hemodialysis secondary to urea compartmentalization.163,164

Single-pool versus double-pool effects in adult and pediatric patients. No single-pool, variable volume model for UKM accounts for the different rates of urea transfer between fluid compartments, commonly termed the double-pool effect. With increased dialyzer efficiency, urea removal from the extracellular compartment (Kp) can exceed its diffusive transfer rate from the intracellular compartment to the extracellular compartment (Kc),165 which is estimated at 800 mL/ min.166 Although higher Kc values have been observed,168,171 the relative impedance of urea movement from the cells to the interstitium and blood effectively render the intracellular compartment an unequilibrated reservoir of urea that is not accounted for by single-pool models of urea kinetics.45,168-170 According to this model (commonly termed the diffusion model), release of the sequestered urea continues for 30 to 60 minutes after completion of the hemodialysis session, a process nonspecifically described as urea rebound.67,119,163,171 Thus, the effective delivered dose of hemodialysis will be overestimated if this sequestered urea pool is large and is not considered (kinetic underestimation of true V).45,164,170,171 The use of dialyzers with high K, especially in patients with a small V, that permit short dialysis times such as pediatric patients (increased K/V), increases the risk of significant double-pool effects.132,133,174

Since the publication of the RPA’s Clinical Practice Guideline on Adequacy of Hemodialysis, another major factor contributing to urea rebound has been characterized: the relative disequilibrium between cardiac output and blood flow to various vascular beds.175 Even if there is no impedance to the movement of urea from the intracellular to the extracellular compartment, organs with relatively low blood flows, such as skin, bone, and muscle, may serve as reservoirs for urea. Seventy percent of the total body water is contained in organs that receive only 20% of the cardiac output (flow-volume disequilibrium). Because these tissues receive relatively less blood flow, their contribution to the total plasma urea content entering the extracorporeal hemodialysis circuit is less than that of tissues that receive more blood flow. Therefore, during the course of a hemodialysis session, flow-volume disequilibrium for urea between vascular beds results in the preferential loss of urea from the well-perfused, but relatively urea-depleted vascular beds. This compartmentalization of urea results in an increase in the BUN concentration over the 60 minutes after the completion of hemodialysis.

The extent of urea rebound varies greatly among adult patients. In one study, the mean amount of urea rebound, measured as the percent increase in postdialysis BUN concentration immediately after dialysis versus 30 minutes after dialysis was 17%.163 However, in some patients, the literature describes the occurrence of as much as 45% rebound171 or 19% to 75% error between single-pool and double-pool KT/V.170 On average, the equilibrated Kt/V (Kt/Vequil) is 0.2 units less than the single-pool Kt/V,176,177 but can be as great as 0.6 units less. For most patients, urea rebound is nearly complete 15 minutes after hemodialysis, but for a minority of patients, it may require up to 50 to 60 minutes.67 As stated above, the severity of urea rebound is partially a function of the K/V.67,163,164,171,174 Therefore, the degree of rebound may be exaggerated in ESRD patients who are small-statured132,170 and during hemodialysis sessions that are complicated by intradialytic hypotension.178

Because of the inability to predict which patients will have significant urea rebound, and in view of the potential deleterious impact of urea rebound on calculations of the delivered dose of hemodialysis,119,132,133,170 the Work Group examined the applicability of the double-pool, variable-volume urea kinetic model. Although this model may more accurately quantify intradialytic urea removal67,170,171 and may result in a more precise NPCR because of the more accurate assessment of V (in comparison to an anthropometric value),169,179 the need to obtain the postdialysis BUN sample 30 to 60 minutes after the completion of hemodialysis (equilibrated postdialysis BUN sample) makes it impractical in the conventional outpatient hemodialysis setting. In an attempt to overcome this practical problem, a mathematical algorithm that estimates the dose of hemodialysis from the predialysis BUN concentration (C0) and the equilibrated postdialysis BUN concentration (Ceq) has been developed (Smye formula).119,133,170 The Ceq is calculated by:

Ceq = C0 × Exp(-[T/(T - TS)] × Ln[CS/Ct])

in which Ceq is the equilibrated postdialysis BUN; C0 the predialysis BUN; CS is the mid-dialysis BUN concentration (usually obtained after 70 minutes of hemodialysis); Ct is the BUN concentration at the end of hemodialysis; T is the duration of hemodialysis; and TS is the time at which CS is drawn. Exp and Ln are the exponential and natural logarithm of the bracketed terms, respectively. The resultant Ceq is used to calculate the Kt/Vequil. In a small study using this formula to generate a calculated equilibrated Kt/V, the average error was only 13%, in comparison to a measured equilibrated value.170

An alternative and more accurate approximation of the equilibrated Kt/V has been developed based on the aforementioned variable organ perfusion, double-pool model (Daugirdas rate formula).174 Assuming all other variables are equivalent, the impact of urea rebound may be much less when using a venovenous angioaccess than with an arteriovenous access (30%)180 (see Guideline 8: Acceptable Methods for BUN Sampling). Thus, the mathematical relationship between the equilibrated and the single-pool Kt/V will vary depending upon the location of the angioaccess and the site from which the postdialysis BUN sample was obtained. Kt/Vsp, calculated using an arterial postdialysis BUN sample from an arteriovenous angioaccess (art Kt/V), will be greater than that calculated from a mixed venous postdialysis BUN, drawn through a venovenous angioaccess (ven Kt/V).181 The corresponding equilibrated Kt/V values are calculated by:

art Kt/Vequil = art Kt/Vsp-(0.6 × art Kt/Vsp/T) + 0.03 (1)

ven Kt/Vequil = ven Kt/Vsp-(0.47 × ven Kt/Vsp/T) + 0.02 (2)

in which T is the dialysis treatment time in hours and t is the dialysis time in minutes.177 The comparative accuracy of the Smye versus the Daugirdas formulations has been determined. In a report from the National Institutes of Health’s Hemodialysis Pilot Study that compared the accuracy of blood side measures of the Kt/Vequil, the rate formula was found to more closely correspond to the equilibrated Kt/V calculated using a 30-minute postdialysis BUN sample (r = 0.85).177 An apparent operational advantage of the Daugirdas formula is that it only requires two blood samples, instead of the three samples for Smye. An alternative, but less well validated two blood sample approach to calculating an equilibrated Kt/V, is to obtain a BUN concentration 30 minutes prior to the completion of dialysis and substitute this value for the postdialysis BUN. This approach is based on the observation that the 30-minute postdialysis BUN concentration is not statistically different than the intradialytic concentration 30 minutes before the end of hemodialysis.182

Using Formula 1 for calculating art Kt/Vequil, the impact of the duration of hemodialysis on the single-pool, variable volume Kt/V is presented in Table I-1. For any Kt/Vsp, shortening the duration of hemodialysis reduces the equilibrated Kt/V. For example, the Kt/V is reduced by >0.18 with a reduction in hemodialysis time from 5.0 to 2.0 hours for all single-pool Kt/V values of 1.0.

Double-pool effects of urea removal can substantially reduce the accuracy of single-pool urea kinetic calculations. Thus, some patients whose dialysis dose is measured by the single pool model alone may be at risk for unappreciated underdelivery of hemodialysis. For this reason, the HD Adequacy Work Group encourages the use of equilibrated Kt/V measurements, especially in patients with a large K/V.183 Use of Kt/Vequil is considerably easier now that the Daugirdas rate formula has been validated.177 The use of two-compartment urea kinetic models to calculate the delivered dose of hemodialysis is a useful adjunct to a complete kinetic modeling program, especially for patients at risk for significant amounts of urea rebound. However, the HD Adequacy Work Group does not yet endorse the substitution of two-compartment urea models for single-pool models in adults on hemodialysis for the following reasons:

1. Two-compartment models based on calculations using the postdialysis BUN sample taken more than 30 minutes after completion of dialysis are impractical in the outpatient hemodialysis setting.

2. Despite theoretical dosing inaccuracy that may result from the failure to take into account the extent of urea rebound, most studies have shown a relationship between doses of hemodialysis estimated by single-pool, variable volume models and patient survival. These studies were the basis of the Work Group’s recommendations about minimal adequate hemodialysis dose.

3. The Work Group was uncertain of the longitudinal validity of either of the mathematical approximations of the equilibrated Kt/V as the patient’s years of survival on hemodialysis increased and/or the patient’s nutritional and cardiac function changed.

4. At this time, there are no prospective outcome studies in adult ESRD patients based on two-pool models. However, at the completion of the HEMO Study, definitive evidence will be provided of the appropriate dialysis dose using equilibrated urea kinetics and will demonstrate its relationship to patients’ mortality and morbidity.92 Similarly, there are no data that address the use of double-pool urea kinetics to predict outcomes in children undergoing hemodialysis.


Table I-1. Relationship Between Single-Pool (spKt/V) and Equilibrated (eKt/V) as a Function of Treatment Time (t)

spKt/V 2.0 hrs 2.5 hrs 3.0 hrs 3.5 hrs 4.0 hrs 4.5 hrs 5.0 hrs
1.0 0.73 0.79 0.83 0.86 0.88 0.90 0.91
1.1 0.80 0.87 0.91 0.94 0.97 0.98 1.00
1.2 0.87 0.94 0.99 1.02 1.05 1.07 1.09
1.3 0.94 1.02 1.07 1.11 1.14 1.16 1.17
1.4 1.01 1.09 1.15 1.19 1.22 1.24 1.26
1.5 1.08 1.17 1.23 1.27 1.31 1.33 1.35
1.6 1.15 1.25 1.31 1.36 1.39 1.42 1.44




1. Data from ongoing prospective, randomized studies, such as the National Institutes of Health HEMO Study,92 that employ double-pool, variable volume models should be used to evaluate the advantages and limitations of double-pool modeling and to clarify its ability to predict patients’ outcomes.

2. The applicability and long-term validity of the formulae that approximate the equilibrated Kt/V need to be determined by observational study.

3. The validity of single-pool, variable volume models and double-pool, variable volume models as outcome predictors needs to be evaluated to a greater extent among pediatric hemodialysis patients.

4. A prospective, multicenter study of the effects of dialysis dose on outcomes in pediatric dialysis patients should be undertaken.

5. Dialysate side and alternative methods of monitoring the delivered dose of hemodialysis, such as online urea measuring devices and direct measures of clearance, need to be developed further and ultimately correlated with patient outcomes in a prospective manner.


Uniformity of Method of Measurement (Opinion)

All patients receiving hemodialysis in the same dialysis facility should have the delivered dose of hemodialysis measured using the same method.

Rationale Adoption of the same method of measurement of hemodialysis dose for all patients in a given facility will enhance consistency in procedures and permit meaningful comparisons of data for a single patient over time between patients and among different hemodialysis facilities. Therefore, the HD Adequacy Work Group strongly urges that the governing body or equivalent administrative authority within the hemodialysis facility formally adopt a uniform method of measuring the delivered dose of hemodialysis for all its patients.

The adoption of such a primary procedure does not preclude the use of supplemental measures of the delivered dose of hemodialysis for some or all patients. For example, some Work Group members felt that monthly URR measurements may be supplemented by less frequent formal UKM. Additional modeling data are certainly not discouraged. However, there should be a single, consistent, and therefore, comparable adequacy measurement for all hemodialysis patients.

If multiple methods of measurement are used within a hemodialysis facility, the method used to quantify the dose of hemodialysis delivered to a given patient should be documented, and the method used to calculate delivered dose of hemodialysis for a given patient should be consistent over time. In the absence of such consistency, longitudinal comparisons of the delivered dose of hemodialysis for a given patient cannot be made.






© 2001 National Kidney Foundation, Inc

web version created by cyberNephrologyTM and The Nephron Information Center