Changing hemodialysis thresholds for optimal survival

Kidney International, Vol. 59 (2001), pp. 738–745

Changing hemodialysis thresholds for optimal survival

LYNDA ANNE SZCZECH, EDMUND G. LOWRIE, ZHENSHENG LI, NANCY L. LEW,J. MICHAEL LAZARUS, and WILLIAM F. OWEN, JR.

Duke Institute for Renal Outcomes Research and Health Policy, Duke Clinical Research Institute, Division of Nephrology,Duke University Medical Center, Durham, North Carolina, and Fresenius Medical Care, North America, Lexington,Massachusetts, USA

Changing hemodialysis thresholds for optimal survival. Numerous studies have demonstrated an associationBackground. The urea reduction ratio (URR), a measure between the amount of hemodialysis and mortality among

quantitating solute removal during hemodialysis, is the frac- patients with end-stage renal disease (ESRD) [1–8]. Intional reduction of the blood urea concentration during a singlegeneral, patient mortality is higher when the hemodialy-hemodialysis treatment. The URR is the principal measure ofsis “dose” is low and is lower when the hemodialysishemodialysis dose in the United States. Based on studies of

patients dialyzed prior to 1994, a minimum URR value of 65% dose is high. The urea reduction ratio (URR) is a com-was recommended to optimize survival. Because of new hemo- monly used measure of hemodialysis dose and is baseddialysis technologies and evolving demographics of the hemo- on the fractional reduction of blood urea nitrogendialysis population, the relationship between the amount of

(BUN) concentration during a single hemodialysis treat-hemodialysis and mortality was examined in contemporarycohorts. ment. It is calculated by dividing the fall of BUN (predial-

Methods. This retrospective cohort included .15,000 patients ysis minus postdialysis BUN) by the predialysis and isper year receiving hemodialysis during 1994 through 1997. Each expressed as a percentage [1, 9–11]. Retrospective stud-patient’s URR was averaged for the three months prior to the

ies of mortality for ESRD patients suggest that the oddsbeginning of each year. Mortality odds ratios were calculatedof death increase progressively as the URR falls belowfor patients by URR. To determine the URR value above

which no further improvement in mortality was seen (“thresh- 60 to 65% [1–3]. Such findings and a professional consen-old”), spline functions were tested in logistic regression models, sus have led three national organizations, including theboth unadjusted and adjusted for case mix measures. The Health Care Financing Administration (HCFA), thestrength of fit for URR, defined by a range of candidate thresh-

principal payer of dialysis services, to advocate a URRolds from 55 to 75%, was evaluated in increments of 1% forof 65% as the threshold for “adequate” hemodialysiseach year using spline functions.

Results. The median URR was 63.2, 65.4, 67.4, and 68.1% and to profile providers accordingly [9, 10, 12–14].for 1994 through 1997, respectively. The median length of hemo- Using URR as a measure of hemodialysis dose, substan-dialysis treatments increased only six minutes from the begin-

tial improvement in the amount of hemodialysis has oc-ning to the end of the period of analysis. Using spline functions,curred in recent years. As reported in the 1998 Nationalthe threshold URR values were 61.1, 65.0, 68.0, and 71.0% for

1994 through 1997 in models adjusted for case mix. The ratio ESRD Core Indicators Report, a profile of patient-of median URR to URR threshold decreased from 1.03 in specific dialysis practices, the mean URR for the United1994 to 0.97 in 1997. States increased from 62.7% (deemed inadequate) toConclusions. From 1994 to 1997, the median URR and the

68.0% (deemed adequate) from 1993 to 1997 [14]. Al-URR threshold for mortality benefit increased. Although anthough mean URR provides a gauge of hemodialysisincreased need in the amount of hemodialysis may be a conse-

quence of changes in patients’ demographic characteristics, the adequacy at the population level, the proportion of pa-likely explanation(s) is a change in the dialysis procedure and/or tients whose URR is $65% is of greater significance forblood sampling favoring higher URR values without changing

optimizing individual patient survival. Using this clinicalthe amount of dialysis provided. The recommended minimumperformance benchmark, the percentage of patients re-URR of 65% appears to be too low to confer an optimal mortal-

ity benefit in the context of current practices. ceiving a URR $65% rose from 43% in 1993 to 72%in 1997 [14].

The aforementioned outcome studies characterizingKey words: end-stage renal disease, dialysis outcome, Medicare, ureareduction ratio, adequate hemodialysis. the association between the amount of hemodialysis and

patient mortality utilized data sets from before 1994Received for publication February 3, 2000[1–8]. However, the applicability of these studies mayand in revised form June 13, 2000

Accepted for publication August 11, 2000 be limited because of evolving demographics of the he-modialysis population and the diffusion of newer hemo- 2001 by the International Society of Nephrology

738

Szczech et al: Hemodialysis dose and optimal survival 739

dialysis technologies. For example, the prevalent hemo- ables, when a single analytic description, formula, or func-dialysis population includes more patients who are older, tion is not valid over the entire range of values. Splinewho have diabetes mellitus, and who are racial and ethnic functions transform these relationships into small seg-minorities [15]. Additionally, prevalent hemodialysis prac- ments of relatively simple functions connected at junc-tices have changed with the increased use of bicarbonate- ture points or “knots.” Initially, two-segment splines werebuffered dialysis solutions, high-flux and biocompatible used to describe the relationship between thresholddialysis membranes, and reprocessed dialyzers [14]. The URR and mortality odds. To identify a threshold, thiseffects of these demographic changes and the diffusion analysis assumes that the relationship between URR andof newer technologies on the relationship between the mortality has two discrete segments. The first segmentamount of hemodialysis and mortality have not been eval- demonstrates an inverse relationship between URR anduated. This study was undertaken to examine the validity

mortality. This inverse relationship continues to a thresh-of the previously demonstrated relationship between

old value of URR, which marks the beginning of the sec-amount of hemodialysis and mortality in the setting ofond segment, above which no further mortality benefitthese secular trends.is observed despite an increasing URR. To assess whichURR value defined the spline with the best fit, URR was

METHODS transformed using each candidate URR threshold byData were taken from the routine analytical files of this equation:

Fresenius Medical Care-North America, Inc. (FMC,Lexington, MA, USA) for the calendar years 1994 (URR 2 t)1 5 50, URR ^ t

URR 2 t, URR , tthrough 1997 [1, 5]. The final sample was comprised ofall patients receiving three times weekly hemodialysis

where the (URR 2 t)1 term evaluates the differencewho were prevalent on January 1 of each year and eitherbetween the measured URR and the candidate thresholdlived the entire year on dialysis or died. Dropouts were(t, from 55 to 75% in 1% increments) for all URR lessnot included in the sample studied. The URR for eachthan each candidate threshold. The “1” subscript indi-patient was calculated by 100 3 (predialysis BUN 2cates that the term has been truncated during the trans-postdialysis BUN) 4 predialysis BUN. Prior to the re-formation process performed in this analysis. All URRlease of the National Kidney Foundation’s Dialysis Out-values $t were assigned a value of 0.comes Quality Initiative on Hemodialysis Adequacy in

For each year, separate logistic regression models esti-1997, no specific protocol was specified for the samplingmated the association between mortality and (URR 2 t)1of blood for the measurement of postdialysis BUN. Fre-using each of the 21 candidate threshold URR valuessenius endorsed a minimum URR or Kt/V of 65% (URR)[19]. This association was modeled in both a univariateor 1.2 [Kt/V (single pool)], respectively, but this was not

mandated. The URR and serum albumin concentrations and multivariable manner. For each year, the strengthfor the last three months of each prior year were aver- of the association between URR and mortality usingaged. All measurements were determined in a single each candidate threshold was estimated in separate logis-laboratory (LifeChem Clinical Laboratories, Rockleigh, tic regression models, unadjusted and adjusted for caseNJ, USA). Only patients with complete demographic mix. The strength of fit for the association between mor-and laboratory data were included in the cohorts. tality and URR as described by each threshold URR

The patients’ demographic and clinical features were was evaluated using the x2 statistic associated with thecompared using the x2 test and the x2 test for trend across (URR 2 t)1 term. The larger the x2 statistic associatedyears. For patients with a URR ,60%, the odds ratios with the spline for each threshold URR, the greater thefor mortality were calculated in URR increments of 5%; amount of variation in outcome explained by URR usingfor patients with a URR $60%, the URR increments

that specific threshold. The x2 statistic for URR as de-were 2.5%, and patients whose URR was .80% com-

scribed by each candidate threshold was plotted againstprised the reference group. Odds ratios were calculatedthe threshold for each year. This analysis was repeatedin univariate and multivariable analyses, the latter con-using Kt/V transformed from URR by taking the nega-trolled for age, gender, race, and diabetes mellitus (casetive loge of one minus the URR divided by 100 [11, 20].mix variables). Risk curves of URR versus log trans-The routine analytical files of FMC for calendar yearsformed odds ratios were plotted for each year.1994 through 1997 do not contain data on the intradia-To describe the relationship between URR and mor-lytic ultrafiltration volume.tality odds, spline functions were used [16, 17], as they

All P values are two-sided, and confidence intervalshave been used in prior analyses of dialysis adequacyare 95%. Analyses were performed using SAS (versionand mortality [2, 18]. Spline functions mathematically

describe complex relationships between continuous vari- 6.08; SAS Institute Inc., Cary, NC, USA).

Szczech et al: Hemodialysis dose and optimal survival740

Table 1. Description of final patient cohorts

1994 1995 1996 1997Characteristic N517,141 N520,785 N515,155 N515,197

Age yearsMean 60.0614.8 60.2615.0 59.5614.9 59.9614.9Median 62.3 62.5 62.0 62.0

Female % 49.3 49.1 49.9 50.0Caucasian race % 49.7 46.1 46.9 50.0Diabetes mellitusa % 39.1 42.4 44.9 47.2Height cm

Mean 168.0610.8 167.7611.0 165.6612.3 165.5612.8Median 167.6 167.6 165.0 165.0

Weight kgMean 70.9617.9 70.9618.1 71.5618.2 72.0618.4Median 68.5 68.3 69.0 69.3

Wt/Ht kg/cmMean 0.4260.10 0.4260.10 0.4360.10 0.4460.11Median 0.41 0.41 0.41 0.42

Albumin g/dLMean 3.8160.42 3.8560.37 3.9060.38 3.8560.38Median 3.83 3.87 3.93 3.88

Dialysis length hoursMeana 3.3860.48 3.3760.49 3.4360.46 3.4860.47Median 3.50 3.50 3.46 3.46

URR %Meana 62.2968.2 65.167.8 66.767.3 67.367.5Median 63.2 65.4 67.4 68.1

Vintage yearsMean (SD) n/a 3.763.8 3.763.6 3.563.6Median n/a 2.5 2.5 2.4

Plus-minus values are means 6 SD; diabetes mellitus describes the presence of diabetes mellitus as a diagnosis or a comorbid condition; vintage refers to thenumber of years since starting dialysis; serum albumin concentrations were drawn predialysis.

a P , 0.0001 for values 1994 to 1997

RESULTS of URR candidate thresholds. Figure 2 plots each URRthreshold and the associated x2 statistic for the case mixThe patients’ demographic features, anthropometricadjusted model. The larger the x2 statistic, the strongerattributes, and distribution of selected laboratory teststhe association between the individual threshold URRare provided in Table 1. Their mean and median agesvalue and mortality. For each year, the curves increasedand composition by gender and race were stable fromas a smooth, monotonic function to a peak and subse-1994 to 1997. Also, anthropometric attributes, describedquently decreased in a similar manner. The URR thresh-by height, weight, and weight/height ratio, were stable.old values with the largest x2 statistic were 61, 65, 68,However, the proportion of patients with diabetes melli-and 71% for 1994, 1995, 1996, and 1997, respectively.tus rose over the period of observation from 39.1% inFor the unadjusted analyses, threshold URR values were1994 to 47.2% in 1997 (P , 0.0001). The mean URR60, 64, 66, and 69%, respectively, for the same years.increased during the period of observation from 62.9%These findings suggest an increase in apparent thresholdin 1994 to 67.3% in 1997 (P , 0.0001). The averagefor hemodialysis adequacy during this period. In addi-length of hemodialysis treatments increased only six min-tional analyses, patients with and without diabetes melli-utes (P , 0.0001). The serum albumin concentration wastus were evaluated separately. For each year, the plotsstable from 1994 to 1997.of URR threshold values and their associated x2 statisticThe association between log odds ratios for mortalityrevealed no consistent differences between diabetics andand URR adjusted for age, race, gender, and diabetesnondiabetics (data not shown).mellitus for each year is presented in Figure 1. During

To assess the effect of different modeled relationshipseach year, as the URR increased, the log odds for mortal-between URR and mortality on the increasing thresholdity decreased. The odds of death declined to an apparentfor adequacy in dialysis dose, complimentary analysesURR threshold, above which increasing URR values didwere performed using two- and three-segment splines.not reduce death risk further. A similar relationship wasOne model assumes that the reduction in death riskrevealed in the absence of adjustment for case mix mea-declines to a nadir, and any additional increase in URRsures. To quantitatively evaluate the perception that theabove this threshold results in a higher mortality (V-URR threshold increased from 1994 to 1997, spline func-

tions were fit in logistic regression models using the range shaped relationship). The second model assumes a re-


Fig. 1. Risk curves for urea reduction ratio(URR) and odds ratio of mortality. The loga-rithm of odds ratio for mortality is plottedagainst URR stratified by year.

verse J-shaped relationship. The initial decline is sepa- 1994, the URR threshold has increased more than themedian URR. This secular trend suggests that improve-rated from the terminal increase in mortality odds by

interposing a flat middle segment, during which no ments in mortality may have not been realized fully.These analyses assume that the relationship betweenchange is seen with increasing URR. Using either as-

sumption, the threshold URR values for each year were URR and mortality has two discrete segments. The firstsegment has an inverse relationship of decreasing mortal-unchanged from the results obtained with two-segment

spline analyses (data not shown). ity with increasing URR, until a threshold URR valueis achieved, which marks the beginning of the secondThe relationship between the median URR and URR

threshold by year is illustrated in Figure 3. Both measure- segment above which no further mortality benefit is ob-tained. Such a relationship has been described by severalments increased from 1994 through 1997. However, the

ratio of the median URR to threshold URR decreased investigators [1, 3, 5, 22, 23]. Alternative models werederived based on descriptions suggesting that the rela-from 1.03 in 1994 to 0.97 in 1997. The slope of this

relationship indicates that the URR threshold increased tionship between URR and mortality may be V or reverseJ shaped [2, 3, 24]. The URR threshold values were notfaster than the median URR.changed by modeling using these different assumptions.This analysis was repeated using Kt/V extrapolated

The increment in URR threshold values was associ-from URR as the measurement of dialysis dose. A similarated temporally with an increase in median URR values.increase in threshold dose was demonstrated over theHypotheses to explain the changing relationship be-period of observation (data not shown).tween hemodialysis dose and mortality include temporalchanges in patient demographics, the hemodialysis pro-

DISCUSSION cedure, and/or prevalent blood sampling techniques. ForThe URR remains the preponderant measure of dial- example, if dissimilar demographic groups have differen-

ysis dose in the United States [14]. This retrospective tial needs for hemodialysis [4, 5], a change in the casestudy of large cohorts of hemodialysis patients demon- mix of prevalent patients may result in an increasedstrates that the conventional minimum URR of 65%, proportion of patients with a need for increased hemodi-recommended to minimize death risk [9, 10, 12–14], may alysis. Some have posited that patients with diabetesbe too low for current hemodialysis practice. The thresh- mellitus require greater doses of hemodialysis to achieveold for the 1994 data set of patients was 61%. Assuming similar survival [4]. While the proportion of patients withvariability of 2.4% for a URR determination [21], a mini- diabetes mellitus in the sample increased between 1994mum URR of 65% was adequate. However, by 1997, and 1997, the URR thresholds for patients with andthe threshold URR value had increased to 71%. This without diabetes mellitus increased over time and werediscordance between the URR value associated with op- not different by diabetic status. Furthermore, the pre-timal mortality benefit (“threshold”), and the conven- ponderance of evidence fails to support a need for in-tional benchmark of 65% suggests that some patients creased hemodialysis for patients with diabetes mellitus

[1, 2, 5].may be placed at excessive death risk. Moreover, since


Fig. 2. Strength of association between URRand mortality. The x2 statistic for each candi-date URR threshold in the logistic regressionmodel of mortality is plotted against the candi-date URR threshold stratified by year.

nutrition and/or inflammation, changed from 1994 to1997 [1, 5, 23, 27, 28].

Another putative explanation to account for the in-creased dialysis dose thresholds is that the URR doesnot measure the removal of the critical uremic toxin. Asmall body of recent evidence suggests that URR doesnot account for the removal of all toxins associated withmortality, such as larger molecular weight solutes [29].Therefore, a higher URR would be needed to achieveremoval of these toxins. This putative mechanism alsoseems unlikely because hemodialyzer technology and thetype of dialyzer used have evolved. Prevalent hemodia-lyzers have better solute removal profiles for large molec-ular weight solutes [14]. If the composition of the cohort

Fig. 3. Temporal trends for the prevalent median URR (x-axis) plotted with significant residual renal function remained theagainst URR threshold (y-axis) values. same during each year, this would introduce another

source of variance into the association between dialysisdose and mortality biasing our ability to determine adifference between years toward the null. Alternatively,Nutrition is a major mortality predictor for ESRDif the composition of the cohort with significant residualpatients, statistically more powerful than the amount ofrenal function decreased in number with each advancinghemodialysis [1, 4, 5, 24]. Recent data suggest that pa-year, the association between dialysis dose and mortalitytients’ anthropometric attributes (weight, body surfacemay be affected shifting the curves to the right. However,area, and body mass index) and the volume of ureaas residual renal function declines with increasing timedistribution (total body water) may be surrogate markerssince the initiation of dialysis, the use of only prevalentfor nutritional status among ESRD patients. Higherpatients minimizes the potential impact of this limitation.measurements are associated with improved survival [5,Supporting our ability to compare the cohorts is that the18, 23, 25]. Lower anthropometric attributes result in alength of time since initiation of dialysis (vintage) hadsmaller urea distribution volume and typically greaternot changed for the years 1995 through 1997.amounts of dialysis [5, 26] but are paradoxically associ-

A systematic change in the dialysis technique and/orated with higher death risks [23]. It is unlikely that thein the timing for drawing blood samples for the URRfindings herein are explained by a trend toward decreas-could also account for these observations. Although theing anthropometric measures among dialysis patients.rate of transfer of urea between intracellular and extra-Neither the cohorts’ anthropometric attributes nor their

serum albumin concentrations, a laboratory surrogate of cellular water is rapid, inequality in urea concentration


may develop between the two compartments during a stimulus for more focused monitoring, profiles, andinvestigations by the regional peer review organizations,rapid hemodialysis. When this occurs, urea that has been

effectively sequestered intracellularly diffuses back into corporate dialysis providers, and individual dialysis units.In the context of substantial pressure to improve dial-the extracellular compartment (“urea rebound”). This

process begins immediately at the end of hemodialysis ysis doses, it is noteworthy that two 1995 surveys ofgreater than 200 hemodialysis centers in the Unitedand lasts over 30 minutes. Therefore, as blood is sampled

over the minutes after the end of hemodialysis, the post- States and Canada reported over 20 different methodsfor sampling blood to measure postdialysis BUN [37, 38].dialysis BUN will rapidly increase, and the calculated

URR becomes smaller. The converse is also true. The As mentioned earlier, a trend toward the use of moreefficient dialyzers has occurred that would favor in-closer the postdialysis BUN is drawn to the end of dial-

ysis, the lower its value and the higher the calculated creased urea rebound [29]. This change in dialyzer typesuperimposed on wide variability in methods for sam-URR. By not accounting for the sequestered amount

of urea, the URR becomes less reflective of total urea pling postdialysis BUN could result in higher prevalentURR values. Moreover, higher URR values achieved inremoval [10]. Secular changes in the hemodialysis tech-

nique, such as the reported increase in the use of higher this or another manner would be welcome by the dialysiscommunity as evidence of betterment. However, in thisflux dialysis membranes that result in increased dialysis

efficiency [14], may cause more urea rebound. If the situation, the URR would be accompanied by a rise inthe threshold for hemodialysis adequacy. If we assumeuse of higher flux membranes is associated with greater

rebound, measured URR would be higher, but the that these findings are the consequence of variability inpostdialysis BUN sampling, this is the first longitudinalamount of hemodialysis delivered would be relatively

little changed. Mortality relationships would shift toward analysis to examine the potential quantitative effect ofthis variability on patient mortality.higher URR values. Higher URR values would compen-

sate for an overestimation of total urea removal. An Recent evidence has suggested that K 3 t is a validoutcome-based measure of hemodialysis dose that is lessalternative, but less frequently used measure of hemodi-

alysis dose, Kt/V, is based on pharmacokinetic theory influenced by anthropometric attributes, which have out-come-associated properties of their own [5, 25]. Amongand attempts to normalize the dose of hemodialysis for

the urea distribution volume [10]. It is unclear whether a patient cohort receiving hemodialysis during 1994, theirmortality risk profile improved until K 3 t was 41.2 tothe use of Kt/V, instead of URR as the predominant

measure of hemodialysis dose, will minimize the appar- 47.3 L per treatment in men and 37.0 to 41.2 L pertreatment in women [25]. There were no additional im-ent need for higher dialysis doses by its improved kinetic

accuracy. This issue may be clarified by the findings from provements in mortality risk for either group as K 3 tincreased above these values. Analysis of a patient co-the National Institute of Diabetes, Digestive, and Kidney

Diseases Hemodialysis Study, which is based on equili- hort receiving dialysis during 1998 demonstrated that aK 3 t of at least 50 L per treatment was required tobrated Kt/V measurements and will be available within

the next two years [30]. Moreover, the randomized, con- achieve maximal death risk reduction [39]. This compari-son suggests that regardless of the measure of dialysistrolled study design will minimize biases inherent to ret-

rospective studies. dose, an increasing standard for dialysis adequacy re-mains evident. Moreover, it underscores that the mea-Finally, substantial pressure has been exerted to im-

prove dialysis outcomes by increasing the amount of surement of dialysis dose rather than the method ofcalculation have evolved.hemodialysis [31–34]. Both URR and Kt/V are calcu-

lated centrally by Fresenius’ laboratory services; the ag- The same is true for dialysis dose calculated as Kt/V.We acknowledge that the transformation formula usedgregate results are reviewed centrally and facility and

patient-specific results reviewed and acted upon locally in this analysis does not account for convective soluteclearance [11], so it is not surprising that the relationshipwithin individual dialysis units. For example, a highly

publicized national discussion proposed linking dialysis between Kt/V and URR thresholds is similar. However,this computational limitation does not compromise thefacility reimbursement to the quality of hemodialysis

services measured using URR [35]. In 1993, a longitudi- relevance of the findings. First, URR remains the princi-pal measure of dialysis dose in the United States [14],nal, nationwide cohort study revealed that only 43% of

the patients had a URR $65% [14]. There resulted an so the findings reported herein are germane to the pre-ponderance of clinical dialysis practices. Second, withinurgency to improve URR values without explicit instruc-

tion regarding strategies to achieve this goal. Recogniz- the FMC system from which the data is derived, andthroughout the United States, few practitioners converting an opportunity and need to improve patient out-

comes, HCFA restructured its method for national URR to Kt/V. Thus, their clinical behavior is driven bythe results reported from URR, rather than the Kt/V.quality improvement in 1994 [14, 36]. Although this was

a population-based initiative, it was intended to serve as Third, to date, there is no peer-reviewed literature that


Reprint requests to Lynda Anne Szczech, M.D., M.S.C.E., Dukedemonstrates that URR is worse than Kt/V as an out-University Medical Center, Institute for Renal Outcomes and Health

come predictor for patient mortality. Thus, although Policy Research, Box 3646, Durham, North Carolina 27710, USA.E-mail: [email protected] Kt/V calculations offer enhanced accuracy for cal-

culating total solute clearance, its relative value overREFERENCESURR is highly debatable for evaluating mortality rela-

tionships. Finally, even if the findings reported herein 1. Owen WF, Lew NL, Liu Y, et al: The urea reduction ratio andserum albumin concentration as predictors of mortality amongwere abrogated by substituting selected Kt/V formulaepatients undergoing hemodialysis. N Engl J Med 329:1001–1006,in the spline models, the URR relationship persists, and1993

it is the URR relationships that are clinically relevant 2. Held PJ, Port FK, Wolfe RA, et al: The dose of hemodialysisand patient mortality. Kidney Int 50:550–556, 1996in the spline models, since this is what most practitioners

3. McClellan WM, Soucie JM, Flanders WD: Mortality in end-use rather than Kt/V.stage renal disease is associated with facility-to-facility differences

In the context of health services, it is worth comment- in adequacy of hemodialysis. J Am Soc Nephrol 9:1940–1947, 19984. Collins AJ, Ma JZ, Umen A, et al: Urea index and other predictorsing about the environment that may have contributed

of hemodialysis patient survival. Am J Kidney Dis 23:272–282, 1994to acceptance of higher URR values, indifferent of how 5. Owen WF, Chertow G, Lazarus JM, et al: The dose of hemodialy-they may be achieved. The renal community’s goal was sis: Mortality responses by race and gender. JAMA 280:1–6, 1998

6. Hakim RM, Breyer J, Ismail N, et al: Effects of dose of dialysisto improve patient health and survival. In this context,on morbidity and morality. Am J Kidney Dis 23:661–669, 1994the mortality rates for the ESRD program nationally 7. Acchiardo SR, Hatten KW, Ruvinsky MJ, et al: Inadequate dial-

and within FMC were increasing from 1991 to 1993, ysis increases gross mortality rate. ASAIO J 38:M282–M285, 19928. Parker T, Husni L, Huang W, et al: Survival of hemodialysissubsequently stabilizing and improving in recent years.

patients in the United States is improved with a greater quantityWhile dose of dialysis is not the only factor that influ- of dialysis. Am J Kidney Dis 23:670–680, 1994ences mortality among ESRD patients [4, 5], the princi- 9. Renal Physicians Association Working Committee on Clinical

Practice Guidelines: Clinical Practice Guideline on Adequacy ofpal focus for affecting this rate became the surrogateHemodialysis. Clinical Practice Guideline, Number 1. Washington,outcome, URR. Concern about the potential for regula- D.C., December, 1993

tory sanctions and an excessive focus of the measurement 10. Hemodialysis Adequacy Work Group: NKF-DOQI clinical prac-tice guidelines for hemodialysis. Am J Kidney Dis 30(Suppl 2):S15–alone may have frustrated adherence to fundamentalS66, 1997principles of continuous quality improvement (CQI). For 11. Lowrie EG, Lew NL: The urea reduction ratio (URR): A simplemethod for evaluating hemodialysis treatment. Contemp Dial Neph-example, Dr. W. Edwards Deming, an internationalrol 11–20, 1991leader in CQI concepts and technology, taught that one

12. American Association of Kidney Patients Advisory: Inadequatemust thoroughly understand the technical and statistical Dialysis Increases the Risk of Premature Death. Tampa, 1995

13. Know Your Number. Baltimore, Health Care Financing Adminis-attributes of a production process for it to be used fortration; Health Standards, Quality Bureau, 1995quality improvement [40, 41]. His principles may be ap-

14. Health Care Financing Administration: 1999 Annual Reportplied to the method by which the URR is measured and End Stage Renal Disease Core Indicators Project. Baltimore, De-

partment of Health, Human Services, Health Care Financing Ad-how changes in the measurement method influence theministration Office of Clinical Standards and Quality, Decemberassociation between the measurement and the clinical 1999

outcome demonstrated here. CQI mandates that if a 15. U.S. Renal Data System: USRDS 1999 Annual Data Report. TheNational Institutes of Health National Institute of Diabetes, Diges-process is altered (change in dialysis membranes and/ortive, Kidney Diseases, Bethesda, April 1999blood sampling techniques) that may modify an essential 16. Shikin EV, Plis AI: Handbook on Splines for the User. Boca

surrogate outcome (URR), its relationship to the ulti- Raton, CRC Press, 199517. Deboor C: A Practical Guide to Splines. New York, Springer andmate outcome must be reevaluated [42]. Until now, the

Herlock, 1978needed validation for the URR-mortality relationship 18. Wolfe RA, Ashby VB, Daugirdas JT, et al: Body size, dose ofhas not been updated. Applying these principles to the hemodialysis and mortality. Am J Kidney Dis 35:80–88, 1999

19. Hosmer DW, Lemeshow S: Applied Logistic Regression. New York,measurement of hemodialysis dose as URR, we need toJohn Wiley & Sons, 1989

recalibrate the definition of hemodialysis “adequacy” 20. Lowrie EG, Chertow G, Lazarus JM, et al: The urea {clearance 3dialysis time} product (Kt) as an outcome-based measure of hemo-and readjust our strategy for achieving the newly defineddialysis dose. Kidney Int 56:729–737, 1999goals. These data suggest the need for uniform blood

21. Kloppenburg WD, Stegeman CA, Hooyschuur M, et al: Assessingsampling practices, a change in the benchmark URR, and dialysis adequacy and dietary intake in the individual hemodialysis

patient. Kidney Int 55:1961–1969, 1999perhaps an increased appreciation of the fundamental22. Gotch FA, Levin NW, Port FK, et al: Clinical outcome relativeprinciples of CQI.

to the dose of dialysis is not what you think: The fallacy of themean. Am J Kidney Dis 30:1–15, 1997

23. Chertow GM, Owen WF, Lazarus JM, et al: Exploring the reverseACKNOWLEDGMENTSJ-shaped curve between urea reduction ratio and mortality. Kidney

We are grateful to Dr. Robert Wolfe, Dr. Frederich Port, and Dr. Int 56:1872–1878, 1999Alan Kliger for helpful discussions during the development of this 24. Lowrie EG, Zhu S, Lew NL: Primary associates of mortalitymanuscript. The views expressed herein are those of the authors and among dialysis patients: Trends and reassessment of Kt/V and ureado not necessarily reflect policy of Fresenius Medical Care, North reduction ratio as outcome-based measures of dialysis dose. Am

J Kidney Dis 32(Suppl):S16–S31, 1998America.


25. Fleischmann E, Teal N, Dudley J, et al: Influence of excess 34. Brink S: Dialysis by the numbers. US News World Report 116:58,1994weight on mortality and hospital stay in 1346 hemodialysis patients.

Kidney Int 55:1560–1567, 1999 35. Congressman Fortney Stark (D-Ca): HR# 1067, 199536. Jencks S, Wilensky G: The Health Care Quality Improvement26. Frankenfield DL, McClellan WM, Helgerson SD, et al: Rela-

tionship between urea reduction ratio, demographic characteris- Initiative: A new approach to quality assurance in medicare. JAMA268:900–903, 1993tics, and body weight for patients in the 1996 National ESRD Core

Indicators Project. Am J Kidney Dis 33:584–591, 1999 37. Beto JA, Bansal VK, Ing TS, et al: Variation in blood samplecollection for determination of hemodialysis adequacy. Am J Kid-27. Owen WF, Lowrie EG: C-reactive protein as an outcome predictor

for maintenance hemodialysis patients. Kidney Int 54:627–636, 1998 ney Dis 31:135–141, 199838. Owen WF, Meyer K, Schmitt G, et al: Methodological limitations28. Yeun JY, Kaysen GA: Factors influencing serum albumin in dial-

ysis patients. Am J Kidney Dis 32(Suppl):S118–S125, 1998 of the ESRD Core Indicators Project: An ESRD Network’s experi-ence with implementing an ESRD quality survey. Am J Kidney29. Leypoldt JK, Cheung AK, Carroll CE, et al: Effect of dialysis

membranes and middle molecule removal on chronic hemodialysis Dis 30:349–355, 199739. Li Z, Lew NL, Lazarus JM, et al: Comparing the urea reductionpatient survival. Am J Kidney Dis 33:349–355, 1999

30. Eknoyan G, Levey AS, Beck GJ, et al: The hemodialysis (HEMO) ratio and the urea product as outcome-based measures of hemodi-alysis dose. Am J Kidney Dis 35:598–605, 2000study: Rationale for selection of interventions. Semin Dial 9:24–33,

1996 40. Deming WE: Out of the Crisis. Cambridge, MIT Centers for Ad-vanced Engineering, 198631. Hull AR: Lessons from Japan can go beyond building automo-

biles. Nephrol News Issues 5:48, 1992 41. Deming WE: The New Economics for Industry, Government Edu-cation (2nd ed). Cambridge, MIT Centers for Advanced Engi-32. Held PJ, Blagg CR, Liska DW, et al: The dose of hemodialysis

according to dialysis prescription in Europe and the United States. neering, 199442. The Food and Drug Administration Modernization Act of 1997Kidney Int 38:S16–S21, 1992

33. Eichenwald K: Death and deficiency in kidney treatment. NY (FDAMA), November 21, 1997. Pub. L. 105–115 b 410(c)(4), 21U.S.C. b 383(c)(4)Times Section A1, 1995

Documents

Changing hemodialysis thresholds for optimal survival