Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (20)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Gutzwiller, J.-P.
Right arrow Articles by Zehnder, C. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gutzwiller, J.-P.
Right arrow Articles by Zehnder, C. E.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Nephrol Dial Transplant (2002) 17: 1037-1044
© 2002 European Renal Association-European Dialysis and Transplant Association

Estimating phosphate removal in haemodialysis: an additional tool to quantify dialysis dose

Jean-Pierre Gutzwiller1,2,, Daniel Schneditz3, Andreas R. Huber2, Christian Schindler4, Felix Gutzwiller5 and Carlos E. Zehnder6

1 Division of Nephrology, Kantonsspital Liestal, University of Basle, Switzerland, 2 Central Laboratory and Division of Nephrology, Kantonsspital Aarau, Switzerland, 3 Institute of Physiology, University of Graz, Austria, 4 Institute for Social and Preventive Medicine, University of Basle, Switzerland, 5 Institute for Social- and Preventive Medicine, University of Zürich, Switzerland and 6 Clinica Las Condes, Santiago, Chile



   Abstract
Top
 Abstract
Introduction
 Subjects and methods
 Results
 Discussion
 References
 
Background. Half of the dialysis population suffers from hyperphosphataemia, which is now recognized as a major factor of haemodialysis (HD) morbidity and mortality. Current control is focussed on reducing dietary phosphate intake and diminishing absorption using phosphate binders, whereas control and quantification of phosphate removal by HD is undervalued. The aim of this prospective study was to develop a simple, bedside formula to estimate dialytic phosphate removal in stable HD patients.

Methods. This was a prospective, randomized trial. Phosphate and urea elimination were assessed in a representative group of patients at two dialysis centres using randomly different dialysers (1.3–2.4 m2). Quantification was performed by partial dialysate collection, concentration measurements in blood and effluent dialysate spot samples, and Kt/Vurea during standard high-flux HD. Multiple linear regression analyses were used in 77% of all data sets to generate an equation to predict phosphate removal. The formula was validated in the remaining 23% of data sets, in the same group of patients using a large capillary filter, and in diabetic patients treated with a small dialyser at different blood flows (200, 250, and 300 ml/min).

Results. A formula allowing quantification of phosphate removal within one HD session was developed in 18 of 74 patients during 41 treatments (137 out of 177 data sets) and was determined as: MPO4pred=0.1t -17+50cds60+11cb60, where t is treatment time in min, cds60 and cb60 are phosphate concentrations in dialysate and plasma measured 60 min into HD in mmol/l, and MPO4pred is estimated phosphate removed in mmol. The precision was remarkable (r2=0.92–0.94). The comparison of phosphate and Kt/Vurea showed a significant association (r2=0.28), albeit with remarkable scatter.

Conclusions. We present the first approach to quantify phosphate removal during high-flux HD by a bedside formula. Only 28% of the variation in phosphate removal was explained by Kt/Vurea. It appears that other factors not adequately accounted for by Kt/Vurea affect phosphate removal. Therefore, we propose an individual control and quantification of phosphate removal in HD.

Keywords: dialysis dose; haemodialysis; hyperphosphataemia; phosphate kinetics



   Introduction
Top
 Abstract
 Introduction
Subjects and methods
 Results
 Discussion
 References
 
Patient morbidity and mortality remains an important issue in haemodialysis (HD). Expected life span is 7–10 years for patients starting HD at the age of 40–44 years, and 4–5 years for those starting dialysis at 60–64 years of age [1]. One of the major mortality risk factors newly emphasized is hyperphosphataemia, which continues to affect about half of the dialysis population, putting patients at significant risk of secondary hyperparathyroidism, vessel and soft tissue calcifications, and death [2,3]. Hyperphosphataemia has also been linked to haemodynamic disturbances such as hypertension, coronary calcifications [4], and left ventricular hypertrophy [5,6], which are discussed to contribute significantly to the high incidence of cardiac death in the dialysis population.

Strategies to control phosphate balance include restriction of phosphate intake determined by diet, reduction of intestinal absorption by phosphate binders, and removal of phosphate by HD. It is very difficult to achieve dietary phosphate restriction in patients with appropriate protein intake. In addition, phosphate-binding agents are not taken regularly. The problem to control phosphate absorption in a clinical setting is given in the following example: phosphate intake with liberal western diet is approximately 25–40 mmol/day (800–1200 mg/day), with 60–86% absorption in the intestine [7]. Therefore, 15–35 mmol phosphate must be eliminated per day, equivalent to 35–82 mmol per thrice weekly HD. A 4-h high-flux HD removes approximately 30 mmol of phosphate per treatment [8]. Each mmol of phosphate in excess of that value requires four capsules (400 mg) of Ca2+-acetate [9], which may require an average of 48 capsules per day. This may lead to serious Ca2+ overload. Surprisingly, control and quantification of phosphate removal by HD is underestimated [10]. However, enhanced phosphate removal by HD must not be dismissed and quantification of phosphate removed by dialysis is mandatory to analyse phosphate balance. Current approaches are cumbersome and the mechanistic structure of phosphate kinetic models is in debate.

Dialysis efficiency is widely assessed by urea kinetic modelling, however, the increase in dialysis efficiency pushed forward by urea kinetics was made without considering possible limitations of other solute transport characteristics within the body. It is likely that an increase in efficiency accompanied by a reduction of treatment time will lead to pitfalls regarding other substances such as phosphate.

Therefore, the aim of this study was to derive a simple statistical model predicting phosphate removal based on a minimum of concentration measurements. The model should be valid for a variety of treatment modalities and a standard dialysis population.



   Subjects and methods
Top
 Abstract
 Introduction
 Subjects and methods
Results
 Discussion
 References
 
This was a prospective, randomized study. It was divided in three parts: A, B, and C. Part A was designed to generate a phosphate model based upon data obtained in high-flux HD using three capillary dialysers with similar surfaces and standard treatment modes. The model was developed using stepwise multiple linear regression analysis considering phosphate removal as dependent variable and markers of potential clinical importance, such as plasma phosphate, effluent dialysate spot phosphate concentrations, Kt/Vurea and dialysis time, as predictor variables. Phosphate mass removal (MPO4) was assessed by partial dialysate collection. Parallel determinations of urea kinetics and urea removal were performed for comparisons. Seventy-seven per cent of the data were used for the model generation. The remaining 23% were used to test the precision of the model.

In part B, the model generated in part A was further validated in a new data-set obtained in the same population using the same treatment mode except for a high-flux dialyser with a larger surface area.

In part C, the model was tested in diabetic patients from another dialysis unit to rule out a centre effect, using the usual dialysate of this centre, a capillary high-flux dialyser with smaller surface area, and different blood flows.

Patients
Patients enrolled in the first two study parts were selected from the HD unit of Aarau. This unit, with 12 000 dialyses a year, is one of the largest in Switzerland and covers about 7% of the Swiss population. Patients enrolled for study part C were selected from the HD unit in Liestal, covering another 5% of the Swiss population. All studied patients were at least one year on HD, stable, and well nourished. Absence of access recirculation was confirmed by monthly screening. Residual renal function was assumed as negligible. Patients gave written informed consent to participate in this study, which was in accordance with the Helsinki Declaration of 1975, as revised in 1983.

Study design
Part A.
Eighteen patients were selected to obtain the data set for developing the phosphate model. Haemodialysis was performed thrice weekly following the standard procedures at the first dialysis unit with an effective blood flow rate corrected for pre-pump blood pressure of 300 ml/min and a dialysate flow of approximately 500 ml/min using two H4008 (Fresenius Medical Care (FMC), Bad Homburg, Germany) dialysis machines. Ultra-pure bicarbonate buffered dialysate with Na+ 140.0, HCO3- 40.0, Mg2+ 0.5 and Ca2+ 1.5 mmol/l was obtained by on-line filtration of mixed dialysate (Diasafe, FMC). Potassium concentration in the dialysate was between 1.0 and 2.0 mmol/l according to plasma potassium concentrations. Dialysate did not contain glucose. Ultrafiltration rates and treatment times were prescribed according to clinical need (4–5 h). Dialyser reuse was not practised.

Data were obtained from a randomized, crossover design using three different high-flux dialysers with similar characteristics. Dialyser I: polyarylethersulfone 2.01 m2 (Arylane H9, Cobe, Hospal, Meyzieu, France); II: polyethylene 2.00 m2 (AM-FP-1000, Asahi Medical Co. Chyoda-ku, Tokyo, Japan); III: polysulfone 1.80 m2 surface area (HF 80, Fresenius AG, Bad Homburg, Germany).

Phosphate and urea kinetic analyses were performed during the mid-week treatment (i.e. on a Wednesday or Thursday). Phosphate and urea removal were measured by partial dialysate and ultrafiltrate collection as described previously [11–15]. The collection system was calibrated before each study. The mean ratio (f=Vd/Vc) of total dialysate volume (Vd) to collected volume (Vc) was 24.9±0.02 (mean±SEM). The total amount of phosphate (MPO4) and urea (Mu) removed was determined by the sum of the solutes in dialysate and ultrafiltrate collected during the first, second, and subsequent treatment hours. In addition, a dialysate spot sample (cds, dialyser outlet) was taken 15, 60, 120, 180, and 240 min into the treatment and at the end of HD. Blood samples were drawn direct from the AV-fistula at the beginning and at the end of HD and from the arterial line in hourly intervals for the measurement of blood urea as well as plasma phosphate. Blood sampling at the end of HD was performed 5 min after the completion of treatment following the recommendations of the NKF-DOQI Clinical Practice Guidelines for Haemodialysis adequacy [16].

Part B.
In this part, the 18 patients were treated and evaluated with the same schedule as in part A, except for a polysulfone dialyser with 2.4 m2 surface area, Dialyser IV (HdF 100S, Fresenius AG, Bad Homburg, Germany).

Part C.
Nine non-obese and stable diabetic patients gave written informed consent to be treated with the same schedule as in part A, except for a dialyser with 1.3 m2 surface area (Dialyser V, F60, Fresenius AG, Bad Homburg, Germany), a dialysate bath with 5.5 mmol/l glucose and 38 mmol/l bicarbonate. They were randomly treated in a crossover design with blood flow rates of 200, 250, and 300 ml/min.

Biochemistry
Plasma was separated by centrifuge within 1 h after collection, phosphate was analysed in plasma and dialysate using an autoanalyser (Dimension RXL, Dade-Behring, Marburg, Germany) according to a modification of Fiske of the classic phosphomolybdate method after precipitation with lithium dodecylphosphate. Urea was measured in heparin-plasma and dialysate according to the urease-glutamate dehydrogenase technique on the Dimension RLX. Coefficients of variation were less than 2.5%.

Calculations
Total dialysate plus ultrafiltrate volume (Vd, in l) was calculated from the volume (Vc) obtained from partial dialysate collections and from the calibration factor (f) determined at the beginning of each study. Ultrafiltrate volume was calculated from patient's weight reductions during dialysis; mass of urea and phosphate removed (Mu, MPO4) was calculated from the mean solute concentration in the collected dialysate and from Vd.

Single pool Kt/Vurea was calculated using the second-generation formula of Daugirdas [17].

Data collected in this way were used for developing an equation by statistical modelling. Since phosphate mass removal was estimated at 2, 3, and 4 h, each 4-h treatment gave three data sets. Dialysis sessions lasting 5 h gave an additional data set.

Model
Stepwise multiple linear regression was performed using independent treatment variables such as plasma and effluent dialysate phosphate concentrations to predict the amount of phosphate removed. The r2 was aimed to be about 0.9 in order to guarantee individual prediction.

Statistics
Data are presented as mean±SD in tables and mean±SEM in figures, respectively. A probability (P) less than 0.05 was considered as significant. Comparison of treatment efficiency in three treatments using different dialysers (I, II, and III) was done by ANOVA for repeated measurements. Binary variables between groups (gender, diabetes) were compared by {chi}2-test. Comparison of predicted to measured phosphate removal was performed according to the method described by Bland and Altman [18]. Analysis was done using Stata v. 6.0 software (Stata Corporation, College Station, Texas, USA).



   Results
Top
 Abstract
 Introduction
 Subjects and methods
 Results
Discussion
 References
 
Eighteen subjects of the population on HD of Aarau (74 patients) participated in study part A and were treated with dialysers I, II, and III, yielding 54 treatments. One treatment was excluded because the vascular access clotted, therefore, 53 treatments were available for the final analysis. In part B, the 18 treatments were evaluated. Nine patients were studied in part C, each with the same filter V at three different blood flows, yielding 27 treatments. One patient missed one study because of acute pancreatitis, three measurements were discontinued because of technical problems and an error in the dialysate bath composition, leaving 23 treatments for final analysis.

Part A—model generation
Physical and treatment characteristics of the study group were representative of the remaining dialysis population (Table 1Go). There was no difference between groups except for a lower frequency for diabetes (28.5% vs 46.6%) and a lower pre-dialysis phosphate concentration (1.44 vs 1.68 mmol/l, P<0.01) in the study group.


View this table:
[in this window]
[in a new window]
  		Table 1.  Characteristics of the study population and the remaining HD population (mean±SD). (Centre A)

 
Treatment characteristics as well as urea (Mu) and phosphate mass removal (MPO4) for dialysers I, II, and III were not different (Table 2Go, Figures 1Go and 2Go).


View this table:
[in this window]
[in a new window]
  		Table 2.  Treatment characteristics of patients dialysed with filter I, II, and III. (Centre A)

 


View larger version (15K):
[in this window]
[in a new window]
  		Fig. 1.  Blood and dialysate urea concentrations in study part A for 5 h dialysis treatments (6 subjects, 18 HDs, 3 treatments, error bars indicate standard error). The dotted line represents the exponential prediction model based on urea measurements. Dialyser I is represented by filled circles, dialyser II by white squares, and dialyser III by filled squares.

 


View larger version (13K):
[in this window]
[in a new window]
  		Fig. 2.  Plasma and dialysate phosphate concentrations in study part A for 5 h dialysis treatments (6 subjects, 18 HDs, 3 treatments). Dialyser I is represented by filled circles, dialyser II by white squares, and dialyser III filled squares.

 
During HD, both urea and phosphate decreased in blood/plasma and in the collected dialysate, however, with different time courses. While urea concentration and urea removal kept decreasing with time by an exponential relationship (Figure 1Go), phosphate concentration rapidly dropped during the first two hours of treatment, but remained stable thereafter (Figure 2Go). Dialytic phosphate removal was highest during the first hour, then it slowed, remaining constant during the last 2–3 h. Hourly urea mass removal continuously decreased from 33±4% during the first hour to 12±2% during the last treatment hour (Figure 1Go). Fractional mass removal of phosphate was 26±6% during the first hour and decreased to a constant level of 18±2, 17±2, and 19±4% during the third, fourth, and fifth hour, respectively, maintaining constant blood concentrations (Figure 2Go). Phosphate elimination improved from 29.8±0.4 to 36.4±0.5 mmol (P<0.0001) by increasing dialysis time from 4 to 5 h.

Phosphate removal improved with increasing Kt/Vurea, but the relationship was unsatisfactory (r2=0.28, P<0.0001) (Figure 3Go). Treatment characteristics for dialysers I, II, or III were not different (Table 2Go).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 3.  Relationship between dose of delivered dialysis (Kt/Vurea) and total mass of phosphate removed per dialysis (MPO4/dialysis) (study part A and B, 18 subjects, 71 treatments, 237 measurements). MPO4/dialysis=12.52xKt/Vurea+7.39, r2=0.28, P<0.0001. The 95% confidence area is defined by the solid lines.

 
Model generation
After stepwise regression, the best model to describe the removal of phosphate as a function of treatment variables was found to be:


(|<|(|>|1|<|)|>|)
where MPO4 is the predicted amount of phosphate removed per dialysis given in mmol/dialysis, time is given in min, cds60 and cb60 are phosphate concentrations in effluent dialysate and plasma measured 60 min into dialysis in mmol/l, respectively.

The result of the regression leading to equation (1) is summarized in Table 3Go. Equation (1) was obtained in approximately 77% (137 out of 177 data sets) of the data sets and was used to predict the phosphate removal in the remaining 23% (40 out of 177 data sets). The predicted values using equation (1) compared to the values measured in the 23% of the data gave a Pearson correlation coefficient of 0.97. Bland–Altman analysis showed an underestimation of 2 mmol/dialysis, which was independent of the amount of phosphate removed (data not shown). The slope of the linear regression was close to unity (MPO4pred=0.92xMPO4meas+0.01, r2=0.94).


View this table:
[in this window]
[in a new window]
  		Table 3.  Regression parameters of the model obtained from 77% of 53 dialysis treatments. (n=137)

 

Part B
To validate equation (1) developed in the first part of the study, 18 treatments were repeated using dialyser IV. Comparison of predicted to the measured phosphate mass removal yielded a Pearson correlation coefficient of 0.97. The slope of the linear regression (MPO4pred=0.91xMPO4meas+0.83, r2=0.93) was close to unity. Bland–Altman analysis revealed a small albeit systematic underestimation of one mmol/dialysis, which was independent of the amount of phosphate removed (Figure 4Go).



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 4.  Bland–Altman plot of phosphate removal measured by dialysate collection and phosphate removal estimated by equation (1). Two standard deviations and the mean differences are indicated by the upper/under and middle lines, respectively (study part B, 18 patients, 18 treatments, 60 measurements).

 

Part C
Equation (1) was further evaluated in high-flux HD with variable blood flows in diabetic patients. Plasma phosphate was higher in this group (Table 4Go). Comparison of the predicted to the measured phosphate mass removal yielded a Person correlation coefficient of 0.94. The slope of the linear regression (MPO4pred=0.94xMPO4meas+2.55, r2=0.88) was close to unity. Bland–Altman analysis revealed a small albeit systematic overestimation of 1 mmol/dialysis, which was independent of the amount of phosphate removed.


View this table:
[in this window]
[in a new window]
  		Table 4.  Patient and treatment characteristics of study part C (9 subjects, 23 treatments, mean±SD). (Centre B)

 
Finally, a correlation of all 104 HD performed for all data sets showed an excellent correlation of 0.92 (Figure 5Go).



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 5.  Relationship between estimated and measured phosphate removal per dialysis in all patients and different dialysers studied (105 HD of 31 studied patients).

 



   Discussion
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
References
 
The purpose of this study was to present a simple formula to predict phosphate mass removal during high-flux HD as an additional and supplementary tool for individual dialysis prescription. The formula is based on treatment time and phosphate concentrations in plasma and effluent dialysate taken at 60 min of the HD session. The equation was developed from 77% of data collected in 53 standard dialysis treatments and successfully tested in the remaining 23%. Further, the precision of this formula was demonstrated using a more efficient dialyser. In spite of this change, the formula was capable to accurately predict phosphate removal. The systematic underestimation of 1 mmol phosphate/dialysis can be assumed as clinically irrelevant, especially because more phosphate is actually removed by this bias. The last part of the study demonstrates that the formula can be applied to diabetics with higher phosphate concentrations, treated with dialysate containing glucose with a smaller high-flux filter and increasing blood flows.

Currently, most dialysis teams prescribe and measure dose of dialysis by urea kinetic modelling using Kt/Vurea. A minimum dose of 1.2 Kt/Vurea units is considered adequate by current standards using a single pool, variable-volume formula [19]. For a given patient volume (V), this target value can be obtained by different means, either by increasing clearance (K) or by increasing time (t). The focus of reaching a target Kt/Vurea of 1.2 units has been on maximizing urea clearance with the side effect of decreasing treatment duration.

It is instructive to compare the dose of dialysis measured as Kt/Vurea in individual treatments with the amount of phosphate removed (Figure 3Go). This comparison is not intended to compare the underlying mechanisms of urea and phosphate removal, which differ in many aspects. At this point it is meant to compare two measures of dose. Intuitively, one would expect an increased removal of phosphate as Kt/Vurea increases. Indeed, there is a relationship between these two variables, but the correlation is poor. Only 28% of the variation in phosphate removal was explained by Kt/Vurea. If a target Kt/Vurea of 1.2 is accepted as a standard, the 95% confidence interval for phosphate removal may cover a range from 15.7 to 29.1 mmol/dialysis. Whereas the same Kt/Vurea may be obtained with increasing blood flows and decreasing treatment times, a decrease in treatment time will substantially reduce overall phosphate removal.

Like other studies [20,21], our data offered evidence of completely different urea and phosphate kinetic profiles. Contrary to urea elimination, phosphate removal is highest in the first one to two hours and then reaches a constant level during the last two to three hours, for a four and five hour HD treatment, respectively. Even though the major fraction of phosphate is removed during the first hour of dialysis, the constant elimination during later stages of the treatment is undervalued. Almost 19% of phosphate is removed during the last hour of a five hour dialysis session compared to only 10% of total urea mass removal. The continuous decrease of urea elimination with time (Figure 1Go) is often used to dismiss an ineffective extension of dialysis duration. This argumentation cannot be applied to phosphate elimination (Figure 2Go). Phosphate removal data indicate that long dialysis sessions significantly contribute to enhanced phosphate elimination (36.4±0.5 mmol/5 h vs 29.8±0.4 mmol/4 h, P<0.0001). The importance of the factor time is documented by time as a linear variable in equation (1). This phenomenon can be explained by the fact that phosphate is mainly distributed in the intracellular space with a slow intra- to extracellular solute transfer rate. Increasing the time of the dialysis session is the best way to raise overall phosphate removal. The clinical importance of treatment time is documented in two recent studies where normal phosphate concentrations could be achieved without phosphate binders using nightly long-term HD [22,23].

High plasma phosphate concentration is a predictor of poor survival on dialysis and in this context, it is believed that hyperphosphataemia is a marker for under-dialysis [24].

Conventional dialysis utilizing high-flux dialysers removes close to 30 mmol phosphate during a 4 h treatment [25]. In order to reduce hyperphosphataemia, with the knowledge that patients' compliance to reduced phosphate diet prescriptions and phosphate binders intake is unsatisfactory, as outlined in the introduction, enhanced removal is required by HD. The current formula offers a practical approach to estimate the dialysis time to obtain a target phosphate removal:


(|<|(|>|2|<|)|>|)

Instead of employing cumbersome dialysate collection, the total amount of phosphate removed can be estimated from two samples and treatment time. Most importantly, samples are taken early in dialysis, soon enough to predict removal as a function of treatment time and to allow for the adjustment of dialysis prescription. The inclusion of phosphate concentration in plasma and dialysate measured at the same time is an indirect measure of dialyser clearance. The plasma value at 60 min (cb60) is a good measure of apparent mean phosphate concentration during the entire treatment, as calculated from the area under the curve and treatment time (AUC/t). The slope of the linear regression between cb60 and AUC/t was close to unity (cb60=1.17xAUC/t-0.12) (data not shown).

A further advantage of this approach is given by the relative independence of dialyser clearance from the specifications given by the manufacturer. As such, the clearance determined by the 60-min concentrations reflects an effective value. This is also evident from results obtained in part B of the study, where a dialyser with increased phosphate clearance was used and precise estimates of phosphate elimination were obtained using equation (1). Intradialytic sampling is not common, but it eliminates the problems associated with post-dialysis rebound. Proper post-dialytic sampling is of special importance when dialysis is quantified by urea kinetic analysis [26].

Recently, two studies demonstrated that hyperphosphataemia is a strong predictor of mortality, independently of Kt/Vurea [22,23].

Kt/Vurea represents a fractional clearance and a dose normalized to initial conditions. It can be used to measure and to prescribe the dose of dialysis. A comparable concept is not available for phosphate kinetics since the exact amount of phosphate to be removed during dialysis is not known from a simple concentration measurement. This question remains to be studied in future. However, the first step into this direction is based on a manageable quantification of phosphate removal. To our knowledge this is the first approach to quantify dialysis dose using phosphate kinetics with a simple bedside formula. The formula developed in this paper explains 88–94% of the variation in phosphate removal in a representative dialysis population treated with high-flux dialysers, dialysate flow of 500 ml/min, and blood flow of 200, 250, and 300 ml/min and is valid for initial plasma phosphate levels between 0.56–2.31 mmol/l. In contrast, the predictive value of Kt/Vurea on phosphate removal was poor. Target values for phosphate removal remain to be identified in future prospective trials.

In conclusion, we call for an individual control and quantification of phosphate removal in dialysis treatment.



   Acknowledgments
 
We wish to thank all the staff of the dialysis units in Aarau and Liestal for their expert technical assistance. We also want to express our special thanks to Karen Noerby and the team of the Central Laboratory, Kantonsspital Aarau. Finally, we are in debt to Professor Werner Zimmerli for useful comments on the manuscript. Part of this study was supported by the Fond für Wissenschaft und Forschung, Kantonsspital Aarau, Switzerland.



   Notes
 
Correspondence and offprint requests to: Jean-Pierre Gutzwiller, MD, Division of Gastroenterology, University Hospital of Basle, CH-4031 Basle, Switzerland. Email: jeanlucjpg{at}hotmail.com Back



   References
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 References
 

  1. United States Renal Data System. USRDS 1998 Annual Data Report. US Department of Health and Human Services. The National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, August 1998
  2. Block GA, Hulbert-Shearon TE, Levin NW, Port FK. Association of serum phosphorus and calciumxphosphate product with mortality risk in chronic haemodialysis patients: a national study. Am J Kidney Dis1998; 31: 607–617[Web of Science][Medline]
  3. Leggat JE Jr, Orzol SM, Hulbert-Shearon TE et al. Noncompliance in haemodialysis: predictors and survival analysis. Am J Kidney Dis1998; 32: 139–145[Web of Science][Medline]
  4. Goodman WG, Goldin J, Kuizon BD et al. Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis. N Engl J Med2000; 342: 1478–1483[Abstract/Free Full Text]
  5. Marchais SJ, Metivier F, Guerin AP, London GM. Association of hyperphosphataemia with haemodynamic disturbances in end-stage renal disease. Nephrol Dial Transplant1999; 14: 2178–2183[Abstract/Free Full Text]
  6. Amann K, Gross ML, London GM, Ritz E. Hyperphosphataemia—a silent killer of patients with renal failure? Nephrol Dial Transplant1999; 14: 2085–2087[Free Full Text]
  7. Ramirez JA, Emmett M, White MG et al. The absorption of dietary phosphorus and calcium in hemodialysis patients. Kidney Int1986; 30: 753–759[Web of Science][Medline]
  8. Zehnder C, Gutzwiller JP, Renggli K. Hemodiafiltration—a new treatment option for hyperphosphataemia in haemodialysis patients. Clin Nephrol1999; 52: 152–159[Web of Science][Medline]
  9. Mai ML, Emmett M, Scheikh MS, Santa Ana CA, Schiller L, Fordtran JS. Calcium acetate, an effective phosphorus binder in patients with renal failure. Kidney Int1989; 36: 690–695[Medline]
  10. Malluche HH, Monier-Faugere MC. Understanding and managing hyperphosphataemia in patients with chronic renal disease. Clin Nephrol1999; 52: 267–277[Web of Science][Medline]
  11. Fellay G, Gabriel JP. Un diviseur de flux qui permet, en cours d'hémodialyse, le receuil continu d'un échantillon représentativ du dialysat. Néphrologie1994; 15: 33–35
  12. Zehnder C, Blumberg A. Influence of dialyser clearance measurement accuracy on haemodialysis prescription based on Kt/V. Nephrol Dial Transplant1994; 9: 753–757[Abstract/Free Full Text]
  13. Charytan C, Gupta B, Meindel N, Spinowitz B. Fractional direct dialysis quantification: a new approach for prescription and monitoring hemodialysis therapy. Kidney Int1996; 50: 1845–1849[Medline]
  14. Blumberg A, Roser HW, Zehnder C, Muller-Brand J. Plasma potassium in patients with terminal renal failure during and after haemodialysis: relationship with dialytic potassium removal and total body potassium. Nephrol Dial Transplant1997; 12: 1629–1634[Abstract/Free Full Text]
  15. Zehnder C, Gutzwiller JP, Huber A, Schindler C, Schneditz D. Low-potassium and glucose-free dialysis maintains urea but enhances potassium removal. Nephrol Dial Transplant2001; 16: 78–84[Abstract/Free Full Text]
  16. NKF DOQI Clinical Practice Guidelines. Blood urea nitrogen (BUN) sampling. Am J Kidney Dis1997; 30 [Suppl 2]: 38–42
  17. Daugirdas JT. Second generation logarithmic estimates of single pool variable volume Kt/V: an analysis of error. J Am Soc Nephrol1993; 4: 1205–1213[Abstract]
  18. Bland MJ, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet1986; 1: 307–310[Web of Science][Medline]
  19. Clinical Practice Guidelines for Haemodialysis Dialysis Adequacy. DOQI Guidelines. Guideline 4: minimum delivered dose of haemodialysis. National Kidney Foundation, 1997
  20. Man NK, Chauveau P, Kuno T, Poignet JL, Yanai M. Phosphate removal during hemodialysis, hemodiafiltration, and hemofiltration. A reappraisal. ASAIO Trans2001; 37: M463–M465
  21. DelSoi CA, Umans JG. Phosphate kinetics during high-flux hemodialysis. J Am Soc Nephrol1993; 4: 1214–1218[Abstract]
  22. Pierratos A, Ouwendyk M, Francoeur R et al. Nocturnal haemodialysis: three-year experience. J Am Soc Nephrol1998; 9: 859–868[Abstract]
  23. Mucsi I, Hercz G, Uldall R, Ouwendyk M, Francoeur R, Pierratos A. Control of serum phosphate without any phosphate binders in patients treated with nocturnal haemodialysis. Kidney Int1998; 53: 1399–1404[Web of Science][Medline]
  24. Ritz E, Schömig M, Bommer J. Osteodystrophy in the millennium. Kidney Int1999; 56 [Suppl 73]: 94–98
  25. Kerr PG, Lo A, Chin MM, Atkins RC. Dialyzer performance in the clinic: comparison of six low-flux membranes. Artif Organs1999; 23: 817–821[Medline]
  26. Owen WF Jr, Meyer KB, Schmidt G, Alfred H. Methodological limitations of the ESRD Core Indications Project: an ESRD network's experience with implementing an ESRD quality survey. Medical Review Board of the ESRD Network of New England. Am J Kidney Dis1997; 30: 349–355[Web of Science][Medline]
Received for publication: 4. 9.01
Accepted in revised form: 22. 1.02


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nephrol Dial TransplantHome page
S. Eloot, W. van Biesen, A. Dhondt, R. de Smet, B. Marescau, P. P. De Deyn, P. Verdonck, and R. Vanholder
Impact of increasing haemodialysis frequency versus haemodialysis duration on removal of urea and guanidino compounds: a kinetic analysis
Nephrol. Dial. Transplant., July 1, 2009; 24(7): 2225 - 2232.
[Abstract] [Full Text] [PDF]

Home page
 Nephrol Dial TransplantHome page
G. Jean, C. Chazot, and B. Charra
Hyperphosphataemia and related mortality
Nephrol. Dial. Transplant., February 1, 2006; 21(2): 273 - 280.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (20)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Gutzwiller, J.-P.
Right arrow Articles by Zehnder, C. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gutzwiller, J.-P.
Right arrow Articles by Zehnder, C. E.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?