Hemodialysis in the Pediatric Patient: State of the Art

Hemodialysis in the Pediatric Patient: State of the Art Stuart L. Goldstein The prevalence of pediatric patients receiving hemodialysis as renal replacement therapy has increased over the past decade. Although numerous technologic advances have been developed and their impact assessed for adult patients receiving hemodialysis, no long-term outcome study currently exists for children receiving hemodialysis. Barriers to such study include the necessity for long-term multicenter participation to enroll enough patients to make definitive statements regarding outcome, lack of consensus for an acceptable and practical method for hemodialysis adequacy measurement in children, and the need for pediatric end-stage renal disease (ESRD)-specific tools for assessment of quality of life. The first part of this article reviews issues surrounding hemodialysis adequacy measurement in children. In particular, simple but accurate Kt/V and normalized protein catabolic rate (nPCR) estimation methods are proposed that should allow for more widespread use of Kt/V and nPCR for measurement of urea clearance and nutritional status in children receiving hemodialysis, important for both patient care and to control for hemodialysis adequacy in pediatric outcome studies. In addition, the principles and pediatric study of 2 technologic advances, continuous noninvasive monitoring of hematocrit and noninvasive ultrasound dilution vascular access flow measurement, are reviewed. Finally, suggestions are provided for future study pertinent to both short-term and long-term outcomes in children receiving hemodialysis. @! 2001 by the National Kidney Foundation, Inc. Index words: hemodialysis, children, adequacy, outcome, ultrafiltration.

he prevalence rate of pediatric patients receiving chronic hemodialysis for mainT tenance renal replacement therapy of endstage renal disease (ESRD) has increased steadily over the last decade. 1 The higher rate might result from numerous factors, including (1) peritoneal membrane failure, precluding the ability to provide adequate peritoneal dialysis; (2) return to dialysis after a failed renal transplant; and (3) a social environment not optimal for home peritoneal dialysis or renal transplantation. Although significant effort has been expended to refine medical and technologic therapies to improve the physical well-being of adult patients with ESRD, no long-term outcome study has been conducted in children receiving hemodialysis. In part, the lack of data stems from 3 facts. ESRD is a relatively rare disorder in children, precluding any single center from caring for enough patients to make definitive statements regarding outcome. Until recently, little work had been performed to validate methods of quantifying dialysis dose in children. Because pediatric outcome studies will require multicenter participation to have sufficient numbers for statistical validity, reliable methods for measuring dialysis adequacy are needed to evaluate

the effect of and control for dialysis dose. Finally, methods to measure health-related quality of life are lacking, and must be developed for assessing outcome of the child with ESRD. The aims of this article are to review recent methods that optimize care of the pediatric hemodialysis patient and to provide a vision for the future role of these methods in assessing and improving the outcome of children receiving hemodialysis.

Hemodialysis Adequacy The concept of hemodialysis adequacy derives from the mechanistic analysis of the National Cooperative Dialysis Study (NCDS) published in 1985.2 A main result inferred from the NCDS was a dramatically higher probability of patient failure (ie, patient death From the Department of Pediatrics, Baylor College of Medicine; and the Renal Service, Texas Children's Hospital, Houston, TX. Address correspondence to Stuart L. Goldstein, MD, Texas Children's Hospital, 6621 Fannin St, MC 3-2482, Houston, Texas 77030. E-mail: [email protected] © 2001 by the National Kidney Foundation, Inc. 1073-4449/01/0803-0004$35.00/0 doi:10.1053/jarr.2001.26347

Advances in Renal Replacement Therapy, Vol 8, No 3 (July), 2001: pp 173-179

173

174

Stuart L. Goldstein

or hospitalization) observed in adult patients who received lower doses of hemodialysis (reported as Kt/V) and/or had worse nutrition (as measured by the normalized protein catabolic rate [nPCR]). Hemodialysis adequacy studies since the NCDS analysis have mainly focused on refinement of Kt/V measuremenp-l0 and determination of the potential effect the hemodialysis dose has on adult patient outcome.1 1-14 Only a few short-term pediatric studies have correlated Kt/V with clinical outcome,15,16 and no published pediatric outcome study has used nPCR as a marker of nutrition status. The NCDS introduced the complex equations of formal urea kinetic modeling (UKM) to calculate Kt/V and nPCR. Treatment data needed for UKM are pretreatment and posttreatment blood urea nitrogen (BUN) levels, dialyzer urea clearance at a particular blood flow rate (K; mL/min), treatment duration (T; minutes), and patient pretreatment and posttreatment weight (kg). Normalized urea clearance during hemodialysis can then be expressed by the unitless term Kt/V. The nPCR (grams of protein/kg/ d) is calculated from the urea generation rate (G) provided by UKM using the modified Borah equation17: 5.43 X G/V d + 0.17. The preference 18-2o for formal UKM for hemodialysis adequacy measurement over other methods resides in its ability to provide a unique solution for both Kt/V and nPCR. Some dialysis providers might be unable or unwilling18 to perform sophisticated UKM because UKM requires advanced computational capabilities19 not readily available to smaller units, including many pediatric centers. Any pediatric hemodialysis outcome study should evaluate and control for both urea clearance and nutrition status, especially because growth and development, which are unique to pediatrics and sensitive to nutritional status, must be included as outcome measures. The evaluation of simpler, surrogate methods to approximate Kt/V has been an active area of investigation. The natural logarithm (In-D) formula of Daugirdas9: nPCR

Kt/V

=

=

-In (Cl/CO

+ (4 - 3.5

0.008 X

X

Cl/CO)

T) X

UF/W

(CO, predialysis BUN; Cl, BUN; T, session duration [hours], UF, ultrafiltration volume (kg); W, postdialysis weight [kg]) had been endorsed 18 as the only valid alternative to formal UKM for Kt/V calculation in adults receiving hemodialysis, because it accounts for both treatment duration and urea removed by ultrafiltration. We compared Kt/V derived by formal UKM with Kt/V derived by the In-D formula in pediatric patients. 3 The correlation of Kt/V derived by each of the 2 methods was extremely high, and the absolute percent error between the 2 methods was less than 6% for every treatment. Partly as a result of this work, the National Kidney Foundation-Dialysis Outcomes Quality Initiative (NKF-DOQI) Hemodialysis Adequacy Guidelines now recommends the In-D formula as a valid alternative to UKM in children. 20 We have recently reported results from a significantly expanded comparison of Kt/V calculated by UKM versus In-D formula and also described a simple method for nPCR calculation. 21 A total of 367 dual adequacy analyses were performed in 39 patients. Kt/V derived by the In-D formula was well correlated with Kt/V derived by UKM and did not vary with patient size (Fig lA,B). For nPCR calculation, an estimated G (estG) was derived using the difference between the posttreatment and pretreatment BUN levels: estG (mg/min)

=

([C2

X

V2] [Cl

X

Vl])/T

(Cl, postdialysis BUN [mg/ dl]; C2, predialysis BUN (mg/ dl); VI, postdialysis total body water [dL; VI = 5.8 dL/kg X postdialysis weight in kg]; V2, predialysis total body water [dL; V2 = 5.8 dL/kg X predialysis weight in kg]; T, time (minutes) from the end of the dialysis treatment to the beginning of the next treatment). Then, estimated nPCR (estnPCR) was calculated using the modified Borah equation (see previously), in which VI represents total body water (L) after dialysis (0.58 X weight in kg). Linear regression analysis showed a high correlation between nPCR derived by each of the 2 methods (Fig lC). Mean difference nPCR was -0.02 + / - 0.04 (range

175

Hemodialysis in the Pediatric Patient

3.0

2.0

.

~

a.

1.5

i

~ 1.0

eZ !l

C

2.5

A

2.0 1.5

~

c: 1.0

0.5

0.5

0.5

1.0

1.5

2 .0

1.0

.

~a.

5

!!

!

~ -5 is If. ·10 ·15

2.0

2.'

3.0

2.5

3.0

0.3

15 10

1 .5

UKM nPCR

UKMlpKW

B

I <20

i! II:

I 20 to 30

I 3Oto4O

I

I

I

~~

~

~

~ :::> 40 to 50

5010150

>to

Wolght(kg)

0.2

D

0.1 Mean + 2SD

,

0.0 Mean

.0.1 Mean - 2S0 .0.2 .0.3 0.0

0.5

. '. 1.0

1.5

2.0

UKMnPCR

Figure 1. (A) Scatterplot with regression line of spKt/V derived by formal UKM versus the natural logarithm formula (r = 0.99, P < .001). (B) Percent error spKt/V plotted against patient weight in 10-kg increment weight subgroups. Mean percent error and standard deviation are depicted by the box and error bars. (C) Scatterplot with regression line of nPCR derived by formal UKM versus the algebraic estimation method (estnPCR) (r = 0.99, P < .001). (D) UKMnPCR versus difference nPCR (UKM nPCR- estnPCR). Only 12 treatments in 9 patients showed a difference in nPCR of greater than 0.1. For all figures, each closed circle represents data from a single patient treatment.

-0.22 to 0.15), and all but 12 treatments in 9 patients had an absolute difference in nPCR of less than 0.1 (Fig 1D). Single-Pool Versus Double-Pool KtIV Kt/V calculation is based on sampling a pretreatment- and posttreatment BUN level. The posthemodialysis BUN concentration rebounds until equilibration occurs 30 to 60 minutes after dialysis. 4,s,22 As the BUN rises after dialysis, the resultant calculation of Kt/V yields lower values. Failure to account for urea rebound causes an overestimation of the true mass of urea removed during dialysis. Calculation of Kt/V by single-pool kinetics (spKt/V) uses the immediate, 3D-second postdialysis BUN (BUN30sec) sample. Calculation of Kt/V by double-pool kinetics (eqKt/V) uses an equilibrated postdialysis BUN. Numerous studies in both children4-s,1O and adults14,22,23 have shown that urea rebound ranges from 7.6% to 24% and accounts for a 12.3% to 16.8% difference between spKt/V

and eqKt/V. Thus, any study evaluating the effect of and/or controlling for dialysis dose on patient outcome should account for urea rebound. Realistically however, it is impractical for patients and potentially costly for a dialysis unit to wait 1 hour after a treatment to obtain an equilibrated BUN (eqBUN) for eqKt/V calculation. As a result, formulas have been devised to estimate eqKt/V by applying a cofactor to spKt/V, relying solely on a pretreatment and a single 30-second-posttreatment BUN level. 6 .7,23 Because urea rebound is primarily characterized by a first-order logarithmic,17 concentration-dependent intracellular fluid (lCF) to extracellular fluid (ECF) urea movement, we hypothesized4,s that eqBUN could be estimated (estBUN) by extrapolating the rise in BUN from 30 seconds to 15 minutes after treatment (LlBUN). Because estBUN is calculated from 2 postdialysis data points, eqKt/V should be more accurately estimated with BUN than with other available I-point formulas. Urea rebound was 69% complete at

176

Stuart L. Goldstein

UF Modeling

22

18 "0

c

:l

.8 a:" ~

=(fJ'1o

14 10

:::l

_2L-________________________- L_ _

o

5

10

15

20

25

30

35

40

45

50

~

55

60

Time after Hemodialysis (minutes)

Figure 2. Logarithmic regression curve derived using 30-second, 5-minute, and 15-minute measured postdialysis BUN values. Urea rebound is 69% complete (REB ls ) at 15 minutes. Reprinted with permission. 4 15 minutes post-dialysis (Fig 2).4 The eqBUN was estimated using the following formula: estBUN = ([LlBUN]/0.69) + BUN30sec The error between estBUN and eqBUN (0.1 to 1.55 mg/ dL) was less than the laboratory measurement standard error for urea nitrogen. The difference between eqKt/V with eqBUN and estimated eqKt/V with estBUN was less than other eqKt/V estimation methods (Table 1).5 Study of online urea monitoring for adequacy calculation in children has shown good correlation with Kt/V by the In-D formula, but only for double-needle hemodialysis. 24 Further study is necessary to assess the practicality of online urea monitoring for adequacy measurement in children.

Accurate determination of patient target weight is an especially challenging task when treating children receiving hemodialysis. The behavior of small children receiving dialysis is often difficult to interpret and some young children are unable to verbalize their symptoms. Month-to-month target weight assessment is difficult because of a widely ranging ratio of total body water to body mass that varies with age. Children are growing and have variable appetites, so real weight loss or weight gain might occur more frequently than in adults. Nonetheless, accurate determination of pediatric target weight is critical, because an underestimation of dry weight can lead to hypovolemia with acute symptoms of hypotension, cramping, and nausea. Chronic overestimation of target weight can lead to chronic volume overload with resultant hypertension, pulmonary edema, congestive heart failure, and left ventricular hypertrophy. A method of continuous monitoring of intravascular volume changes has been developed,25 based on the principle that because red cell volume remains constant during dialysis,26 changes in hematocrit will be inversely proportional to changes in intravascular volume. Continuous optical methods of noninvasive monitoring of hematocrit (NIVM) take advantage of this relationship to show a realtime association between fluctuating hematocrit and intravascular volume during the HD treatment. Theoretically, the use of optical methods of NIVM to continuously monitor changes in intravascular volume can help decrease clinical symptomatology during a dialysis treat-

Table 1. Comparison Between Measured eqKt/V and Estimated eqKt/V Method

eqKt/V* estKt/V (Goldstein) estKt/V (Tattersall) estKt/V (Daugirdas) estKt/V (Maduell)

Mean Kt/V

1.18 ± 1.19 ± 1.19 ± 1.17 ± 1.14 ±

Mean Absolute Percent

0.20 0.21 0.18 0.17 0.20

* Kt/V calculated with measured 60-minute post-BUN (eqBUN).

LlKt/Vt

NA 3.4% ± 2.3% 4.5% ± 3.9% 4.4% ± 3.7% 6.7% ± 8.3%

t The absolute percent difference between the particular estimated Kt/V and the eqKt/V. :j: Values for total percent error = mean absolute percent error + 2SD. Reprinted with permission.s

Mean Total Percent LlKt/V:J:

NA 8.0% 12.3% 11.8% 23.3%

Hemodialysis in the Pediatric Patient

ment. For example, if fluid removal is too rapid, as detected by a rapidly increasing hematocrit and steep declining slope on the NIVM monitor, the UF rate could be decreased before hypovolemic symptoms occur in the child. If fluid removal is too slow, as suggested by an unchanging hematocrit and a flat slope on the NIVM monitor, the UF rate may be increased safely to achieve the patient's actual target weight. Steuer27 achieved a 2-fold reduction in intra dialytic symptoms with NIVM in adult patients prone to hypotension without altering treatment times or volume removed. We studied the effect of NIVM on intradialytic symptomatology in pediatric patients receiving hemodialysis. 28 A UF-associated event was defined as a patient symptom (restricted in the study to hypotension, headache, or cramping) that required nursing intervention (saline bolus, Trendelenburg position, slowing of or cessation of UF). The event rate (number of events per number of treatments) was lower, especially for patients less than 35 kg, when NIVM was performed. Decrease in event rates with NIVM occurred despite both a lower nurse-to-patient ratio and aggregate nursing experience and without a sacrifice in target weight achievement. When NIVM was performed, events in the first 90 minutes of a treatment occurred only with a blood volume change on the NIVM monitor of greater than 8% per hour. A total of 71% of events occurring after 60 minutes of treatment initiation were associated with an hourly blood volume change greater than 4%. The results of this study suggest that UF might be modeled effectively using NIVM to optimize fluid removal and minimize patient symptoms. For instance, NIVM can guide provision of higher UF rates early in the treatment followed by lower UF rates in the last hour of treatment. In addition, NIVM has been an invaluable tool to help differentiate true UF-associated symptomatology from patient anxiety or fear in the younger patients.

Noninvasive Permanent Hemodialysis Vascular Access Monitoring Provision of adequate hemodialysis depends upon a properly functioning vascular access.

177

A permanent vascular access in the form of an arteriovenous fistula (AVF) or graft (AVG) can function for many years and are preferred over indwelling catheters in children. Thrombosis of a permanent access is a significant cause of morbidity for the hemodialysis patient population. 1 In many instances, thrombosis occurs as a result of decreased access flow caused by a stenosis of the access venous outflow tract. Ultrasound dilution (UD), a practical, noninvasive, and reliable indicator of vascular access flow, has been used effectively to identify venous stenosis in adult patients receIvmg hemodialysis. 29 -31 UD employs 2 reusable sensors, each attached to temporarily reversed venous and arterial lines. A 20-mL bolus of saline is injected quickly into the venous line proximal to the venous sensor. The sensors are attached to a computer that interprets the changes in Doppler velocity within each line as the hematocrit changes in relation to dialyzer blood flow. In adults, a 6-month decrease 30 in access flow of greater than 15% or an absolute access flow less than 650 mL/min31 as measured has been associated with an increased risk of access thrombosis or stenosis. We evaluated the accuracy of UD in predicting AVG and AVF stenosis in pediatric patients receiving hemodialysis. 32 When access flow (QA; mL/min) reported by the UD monitor was corrected for patient size (QAcorr, mL/min/1.73 m 2 ), QAcorr was significantly lower in accesses with stenoses (mean, 401 + 176 mL/min/1.73 m 2; range, 174 to 579) versus accesses without stenosis (1158 + 330 mL/min/1.73 m 2 ; range, 709 to 1711). No patient with an QAcorr score of greater than 700 mL/min/1.73 m 2 developed thrombosis within 30 days after UD measurement and 2 patients with a QAcorr of less than 600 mL/ min/1.73 m 2 developed thrombosis within 1 week after measurement. These data suggest that a QAcorr of less than 600 mL/min/1.73 m 2 might be predictive of severe vascular access (VA) stenosis and should lead to prompt referral for angioplasty to prevent imminent access thrombosis. Other noninvasive VA flow measurements include ionic dialysance (ID) and transcutaneous access flow (TQA). The ID model infers blood flow from the difference in solute con-

178

Stuart L. Goldstein

centration predialyzer and postdialyzer in both normal and reversed line configurations. MercadaP3 showed a good correlation in adults between QA measured by UD and ro. TQA uses a disposable probe placed directly over the patient access and connected to an NIVM monitor. A saline bolus is injected in the arterial line. The TQA probe measures the transient drop in hematocrit in the access. Studies in adults show that TQA correlates well with UD.34 Studies have yet to be conducted in children with these methods.

Vision for the Future of Pediatric Hemodialysis Evaluation and refinement of methods addressed in this article have been important steps toward establishing an evidence-based standard of optimal care for children receiving hemodialysis. Such a standard must be based on assessment of both the acute and chronic effects of hemodialysis on short- and long-term outcomes. Thus, relevant outcome measures for children with ESRD must be determined. Potential areas for investigation include evaluation of the effect of different UF modeling algorithms on postdialysis thirst, fatigue, interdialytic weight gain, exercise capacity, mental function, and patient attitude toward dialysis. Daily home hemodialysis improves blood pressure control, appetite, phosphate control, and quality of life in adult patients3S and would likely improve these parameters for children and adolescents who could safely perform the procedure. In addition, establishing pediatric vascular access monitoring standards with UD or other noninvasive techniques will likely prolong vascular access survival rates and decrease patient morbidity and hospitalization. Such studies can only help to optimize patient physical and psychological well-being. Long-term outcome studies will need to be multicenter in design and will likely follow patients from hemodialysis through transplantation. Some preliminary single-center work has already shown the value of controlling for adequacy measurement in terms of short-term pediatric patient outcomes.1s It has not been the goal of our center's investigation on adequacy to define a minimally acceptable

level of urea clearance. However, these adequacy measurement methods allow for clinically applicable adequacy assessments and should serve well to control for Kt/V and nPCR to determine their potential impact on pediatric outcomes. Further quests for more accurate measures of adequacy might be not be necessary and could be fruitless, given the simple and reliable equations validated in our research and the inherent error in laboratory BUN measurement. s Whereas methods to optimize dialysis care have been refined, no practical tools specific to pediatric patients with ESRD exist that longitudinally assess health-related quality of life issues such as fatigue level, exercise capacity, medication adherence, and barriers to advanced educational and vocational training. Such tools must be developed and validated to evaluate the effect of numerous factors, including Kt/V, nPCR, anemia, and socioeconomic status on the long-term outcome for pediatric patients with ESRD.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

us Renal Data System: Excerpts From the USRDS 2000 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. Am J Kidney Dis 36:S1S238, 2000 (suppl 2) Gotch FA, Sargent JA: A mechanistic analysis of the National Cooperative Dialysis Study (NCDS). Kidney Int 28:526-534, 1985 Goldstein SL, Sorof JM, Brewer ED: Naturallogarithmic estimates of Kt/V in the pediatric hemodialysis population. Am J Kidney Dis 33-518-522, 1999 Goldstein SL, Sorof JM, Brewer ED: Evaluation and prediction of urea rebound and equilibrated Kt/V in the pediatric hemodialysis population. Am J Kidney Dis 34:49-54, 1999 Goldstein SL, Brewer ED: Logarithmic extrapolation of a 15-minute postdialysis BUN to predict equilibrated BUN and calculate double-pool Kt/V in the pediatric hemodialysis population. Am J Kidney Dis 36:98-104,2000 Marsenic OD, Pavlicic D, Peco-Antic A, et al: Prediction of equilibrated urea in children on chronic hemodialysis. ASAIO J 46:283-7, 2000 Sharma A, Espinosa P, Bell L, et al: Multicompartment urea kinetics in well-dialyzed children. Kidney Int 58:2138-2146, 2000 Evans JH, Smye SW, Brocklebank JT: Mathematical modelling of haemodialysis in children. Pediatr Nephrol 6:349-353, 1992 Daugirdas JT: Second generation logarithmic estimates of single-pool variable volume Kt/V: An analysis of error. J Am Soc Nephrol4:1205-1213, 1993

Hemodialysis in the Pediatric Patient

10. Tattersall J., DeTakats D, Chamney P, et al: The posthemodialysis rebound: Predicting and quantifying its effect on Kt/V. Kidney Int 50:2094-2102, 1996 11. Held PJ, Port FK, Wolfe RA, et al: The dose of hemodialysis and patient mortality. Kidney Int 50:550-556, 1996 12. Bloembergen WE, Stannard DC, Port FK, et al: Relationship of dose of hemodialysis and cause-specific mortality. Kidney Int 50:557-565, 1996 13. Hakim RM, Breyer J, Ismail N, et al: Effects of dose of dialysis on morbidity and mortality. Am J Kidney Dis 23:661-669, 1994 14. Daugirdas JT, Depner TA, Gotch FA, et al: Comparison of methods to predict equilibrated Kt/V in the HEMO Pilot Study. Kidney Int 52:1395-1405,1997 15. Tom A, McCauley L, Bell L, et al: Growth during maintenance hemodialysis: Impact of enhanced nutrition and clearance. J Pediatr 134:464-471, 1999 16. Brem AS, Lambert C, Hill C, et al: Outcome data on pediatric dialysis patients from the end-stage renal disease clinical indicators project. Am J Kidney Dis 36:310-317, 2000 17. Borah MF, Schoenfeld PY, Gotch FA, et al: Nitrogen balance during intermittent dialysis therapy of uremia. Kidney Int 14:491-500, 1978 18. National Kidney Foundation-Dialysis Outcomes Initiatives Clinical Practice Guidelines. Am J Kid Dis 30:S1-S62, 1997 (suppl 2) 19. Depner TA: Single compartment model, in Depner TA (ed): Prescribing Hemodialysis: A Guide to Urea Modeling. Boston, MA, Kluwer, 1991, pp 65-89 20. National Kidney Foundation-Dialysis Quality Outcomes Initiatives Clinical Practice Guidelines. Am J Kid Dis 37:S7-S64, 2000 (suppl 1) 21. Goldstein SL, Brewer ED: Single-pool Kt/ V and nPCR are both reliably estimated using simple formulas in pediatric hemodialysis (HD) patients. J Am Soc Neph 11:320, 2000 (abstr) 22. Pedrini LA, Zereik 5, Rasmy S: Causes, kinetics and clinical implications of post-hemodialysis urea rebound. Kidney Int 34:817-824, 1988 23. Daugirdas JT, Schneditz D: Overestimation of hemo-

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

179

dialysis dose depends on dialysis efficiency by regional blood flow but not by conventional two pool urea kinetic analysis. ASAIO J 41 :M719-24, 1995 Van Hoeck KJ, Lilien MR, Brinkman DC, et al: Comparing a urea kinetic monitor with Daugirdas formula and dietary records in children. Pediatr Nephrol 14: 280-283,2000 Steuer RR, Leypoldt JK, Cheung AK, et al: Hematocrit is an indicator of blood volume and a predictor of intradialytic morbid events. ASAlO J 40:M691-M696, 1994 Swartz RD, Somermeyer MG, Hsu C-H: Preservation of plasma volume during hemodialysis depends on the dialysate osmolality. Am J Nephrology 2:189-194, 1982 Steuer RR, Leypoldt JK, Cheung AK, et al: Reducing symptoms during hemodialysis by continuously monitoring the hematocrit. Am J Kidney Dis 27:525532, 1996 Jain SR, Smith L, Brewer ED, et al: Non-invasive intravascular monitoring in the pediatric hemodialysis population. Ped Nephrol 16:15-18, 2001 Krivitski NM: Theory and validation of access flow measurement by dilution technique during hemodialysis. Kidney Int 48:244-250, 1995 Neyra NR, Ikizler TA, May RE, et al: Change in access blood flow over time predicts vascular access thrombosis. Kidney Int 54:1714-1719, 1999 May RE, Himmelfarb J, Yenicesu M, et al: Predictive measures of vascular access thrombosis: A prospective study. Kidney Int 52:1656-1662, 1997 Goldstein SL, Allsteadt A: Ultrasound dilution evaluation of pediatric hemodialysis vascular access. Kidney Int 59:2352-2360, 2001 Mercadal L, Hamani A, Bene B, et al: Determination of access blood flow from ionic dialysance: Theory and validation. Kidney Int 56:1560-1565, 1999 Steuer R, Miller, L, Zhang, B, et aI: Non-invasive, transdermal access blood flow (TQa) . JAm Soc Neph 11:300,2000 (abstr) Pierratos A: Daily hemodialysis: Why the renewed interest? Am J Kidney Dis 32:576-S82, 1998 (suppl 4)