Secondary Logo

Journal Logo

Dialysis & Kinetics

Online Conductivity Monitoring: Validation and Usefulness in a Clinical Trial of Reduced Dialysate Conductivity

Lambie, Stewart H.; Taal, Maarten W.; Fluck, Richard J.; McIntyre, Christopher W.

Author Information
doi: 10.1097/01.MAT.0000150525.96413.AW
  • Free


The mortality and morbidity of chronic hemodialysis patients are critically dependent upon the maintenance of appropriate salt and water balance. Both excessive and insufficient removal of the interdialytic sodium load can lead to significant problems. These problems are manifested either as hemodynamic instability or as a state of salt and water overload. The major influence over the total mass of sodium removed is the dialysate sodium concentration ([Na+]). Setting the dialysate [Na+] has largely been an expedient response to intradialytic symptoms and hemodynamic instability.1–4 Before the development of blood pumps and negative pressure dialysis systems (allowing significant mechanical ultrafiltration), hypotonic dialysate was used to enhance sodium and water removal.5 Use of such low dialysate [Na+] was associated with high levels of dialysis discomfort6 and the potentially fatal disequilibration syndrome.7 The past 20 years or so have seen a trend towards reduction of hemodialysis treatment times. This has aggravated the reliance on “high” sodium dialysate to allow subjectively acceptable treatment delivery.

Unfortunately, whereas increased dialysate [Na+] provides more hemodynamic stability during dialysis, increased sodium intake from dialysate or diet is associated with increased interdialytic weight gains,8 hypertension,9 increasing left ventricular mass,10 and potentially worsening clinical outcomes.11,12 Lower dialysate [Na+] levels would, therefore, appear to be desirable in the longer term, if this can be achieved without unacceptable intradialytic hemodynamic instability.

At present, common clinical practice is to prescribe a single default dialysate [Na+] for the entire hemodialysis population. Patients on hemodialysis vary in their degree of adherence to suggested dietary sodium and fluid restrictions,13 which suggests that, in addition to prescribing individualized amounts of fluid removal, we should also prescribe individualized amounts of sodium removal. Predialysis plasma [Na+] varies between individuals but is relatively constant over time within individuals.14 Thus a single dialysate [Na+] for all patients will lead to varying amounts of sodium removal. Sodium will diffuse down the concentration gradient between the plasma and dialysate [Na+], leading to variable diffusive sodium removal depending upon the predialysis plasma [Na+]. An individualized approach to dialysate [Na+] has been recommended previously,14,15 and one study exploring some aspects of this concept has been reported.16

Recently, dialysis technology has been developed, based upon the original theory developed by Polaschegg,17 with the capability of the routine prescription of an individualized dialysate sodium based upon a computerized dialysis control system. The use of online monitoring of plasma conductivity (Cp) by the measurement of ionic dialysance has demonstrated that such technology is capable of measuring sodium flux during hemodialysis.16,18,19 The combined use of these technologies can achieve individualized dialysate sodium, as well as the ability to closely monitor the resulting changes in sodium balance.

The current study aims to reinforce the previous validation of online conductivity monitoring in measuring plasma sodium and sodium flux and to examine the clinical utility of conductivity monitoring in a randomized trial of the empirical reduction of dialysate conductivity (Cndi). This aspect of the study will assess the clinical impact and practicality of Cndi reduction and will use online conductivity monitoring to follow the resultant changes in sodium balance.

Methods and Materials

A total of 28 patients on chronic hemodialysis were randomized into one of two groups. The first group underwent progressive empirical reduction of Cndi in steps of 0.2 mS/cm. If they remained hemodynamically stable, they were maintained on this prescription for at least 2 weeks, and then the process was repeated until a prespecified minimum conductivity of 13.0 mS/cm was achieved. This extremely low level of conductivity (equivalent to 130 mmol/L Na+) was set to test the hypothesis that radical reduction is possible, at least in the short term, and to study sodium dynamics over as wide a range as possible. Patients with excessive weight gains or hemodynamic instability on dialysis were excluded.

Hemodynamic instability was assessed after every dialysis session. The markers used were as follows: (1) symptoms of hypotension as noted by nursing staff, (2) relative blood volume (RBV) falling below 10%, (3) systolic blood pressure (BP) falling below 90 mm Hg, and (4) administration of saline. Patients were returned to the previous Cndi if one of these markers had occurred during the preceding week. The second group continued to dialyze against Cndi of 13.6 mS/cm (equivalent to 135 mmol/L Na+) throughout the study.


Mean age of the recruited patients was 57 ± 3 (50–64 years), mean time on dialysis was 47 ± 11 (23–71) months, and mean dry weight was 77 ± 2.5 (73–83) kg; 13% were diabetic, 33% had ischemic heart disease, and 66% were taking no antihypertensive medication. Patients gave appropriate informed written consent for participation in the study, which had been approved by the Local Regional Ethics Committee.

Hemodialysis Schedule

HD was performed using Hospal Integra dialysis monitors (Hospal, Mirandola, Italy) equipped with Diascan conductivity monitoring modules. All machines were also equipped with a blood volume monitoring capability based upon hemoglobin concentration measurements (Hemoscan). All patients dialyzed for 4 hour treatment sessions, three times per week. Patients were dialyzed using hemophan dialyzers (Hospal HG 500–700) and bicarbonate buffering. The other components of the dialysate were as follows: −K+ mmol/L, Ca++ 1.25 mmol/L, Glu 5.6 mmol/L, Mg++ 0.5 mmol/L, Cl 107.5 mmol/L, HCO3 32 mmol/L, and CH3COO 3.0 mmol/L.

Dialysis prescriptions were held on a separate server and downloaded to the dialysis monitor for each treatment. All data pertaining to the dialysis session were uploaded to patient specific files at the end of each treatment for subsequent analysis. No changes to dry weight, nor to any aspect of dialysis prescription except for Cndi, were made during the course of the study. Kt/V was not specifically measured.

Assessments of Sodium Flux

Measurement of sodium loading was made by noninvasive, online conductivity monitoring using the Diascan module and ionic dialysance measurements. This system possesses conductivity probes at both the dialysate inlet and waste dialysate outlet. Cndi is transiently increased to a known value intermittently throughout the hemodialysis session (each 30 minutes). Measurement of the changes in conductivity at the dialysate waste outlet allows ionic dialysance to be measured and effective Cp to be derived (and continuously displayed on the dialysis monitor).

The Diascan module also derives a value for ionic mass balance (IMB) as a measure of total ionic flux across the dialyser. This is predominately a consequence of sodium movements20 and has been validated against conductivity measurements of direct dialysate quantification.16 Positive values for IMB indicate sodium removal from the patient.


We further validated the Diascan results with direct measurements of pre- and postdialysis plasma [Na+] using an indirect ion selective electrode based method (assay CV 0.58–1%). These were compared with initial and postdialysis Cp. Similarly, pre- and postdialysis plasma [Na+] were used to derive sodium loss. For this calculation, we assumed that the volume of distribution of sodium was equivalent to total body water (TBW) and that postdialysis is estimated as being 58% of target weight,21 whereas predialysis is estimated as being 58% of target weight plus 100% of ultrafiltration volume. Hence

The assumption that sodium behaves as though distributed in TBW is uncertain, and the true volume of distribution may be less than this, but, in the absence of a reliable tool with which to estimate extracellular fluid, it is reasonable to use this approximation.

The convective contribution to total IMB (IMBtotal) was calculated using the mean of initial and final Cp for each treatment to estimate the mean plasma [Na+] during dialysis and, therefore, the mean [Na+] in the ultrafiltrate. A fixed ultrafiltration rate was used without ultrafiltration profiling. For the calculation of convective sodium loss, conductivity was converted to [Na+] using the following relationship:22

Mean [Na+] was then multiplied by the ultrafiltration volume to derive the convective sodium loss (IMBconv), and diffusive IMB (IMBDiff) was calculated by subtracting convective from total IMB. Thus the equation for IMBDiff is as follows:

To confirm that there was a linear reduction in Cp over the course of a dialysis session, and hence that using the arithmetic mean of Cp in the previous calculation is a reasonable approximation to the mean conductivity of ultrafiltrate, the reductions in Cp at 30 minute intervals for 30 treatments were studied.

Symptom Scoring

Scores for thirst and cramps were recorded using a visual analog scale at baseline, at lowest Cndi, and again after reverting to baseline. A score of 10 represents extreme thirst or cramp, and 0 represents no symptoms.

Statistical Analysis

All data were analyzed using GraphPad Prism version 3.00 for Windows (GraphPad Software, San Diego, CA). Correlation plots were subsequently analyzed by linear regression. Coefficient of determination was calculated from the Pearson correlation coefficient. Data are expressed as mean ± sem (95% confidence intervals) unless otherwise stated. Box and whisker plots display median and range. The box extends from the 25th to the 75th quartile. Normally distributed paired data were analyzed by paired t-test, nonparametric paired data by Wilcoxon matched pairs test, and nonpaired, nonparametric data by Mann-Whitney test.



In the validation phase, 57 pre- and postdialysis samples were available from 20 patients. Plasma [Na+] correlated with Cp (R2 = 0.61, p < 0.0001). The relationship between plasma [Na+] and Cp giving the best fit with the data was as follows: Plasma [Na+] = [13.36 × Cp] − 49.18. Sodium removal estimated from pre- and postplasma [Na+] correlated well with sodium removal measured by IMB (R2 0.66, p < 0.0001). IMB underestimated sodium loss by 14.8 ± 3.3 (8.2 to 21.3)% (Figure 1), and this did not appear to vary throughout the range of sodium loss studied.

Figure 1.
Figure 1.:
Bland-Altmann plot of the difference between sodium removal by plasma concentration of sodium and IMB, expressed as a percentage of the mean, and plotted against the mean of the two measurements, demonstrating good agreement between the two measures throughout the range of values obtained. Percentage difference 14.7 ± 3.3 (8–21)%. IMB, ionic mass balance.

There was a linear fall in Cp during the 30 treatments studied with repeated half hourly measurements (mean R2 0.89, range 0.84–0.93) (Figure 2). Use of mean initial and final Cp to estimate time averaged Cp during dialysis is, therefore, reasonable as is used in the calculation of IMBDiff.

Figure 2.
Figure 2.:
Cp over time in five representative patients during a single dialysis session. Results show a linear decline in Cp during hemodialysis (R2 = 0.89, range 0.84–0.93). Cp, plasma conductivity.

Intervention Group

Clinical results.

In the group randomized to adjustment of Cndi, all 16 patients tolerated the initial reduction to 13.4 mS/cm, 12 tolerated further reduction to 13.2 mS/cm, and 6 tolerated the final reduction to 13.0 mS/cm. Reduction of Cndi was predominantly limited by symptomatic intradialytic hypotension, which occurred in 80% of the sessions that demonstrated hemodynamic instability. There was no difference between these groups in age, weight, time on dialysis, comorbidity (ischemic heart disease and diabetes mellitus), and number of antihypertensive medications. Baseline measurements, taken during the initial sessions at Cndi 13.6mS/cm, were also analyzed for the three groups achieving different conductivities and were only significantly different for final RBV, comparing 13.4 group with 13.0 group. The 13.0 group achieved greater RBV reduction in their baseline dialysis sessions (Table 1). Data were initially analyzed comparing all recorded dialysis sessions at differing Cndi.

Table 1
Table 1:
Baseline measurements at Cndi 13.6 mS/cm for groups 1, 2 and 3.

Interdialytic weight gain showed a significant progressive reduction for sessions with lower Cndi (Figure 3). Mean interdialytic weight gain was 1.40 ± 0.12 kg for Cndi 13.0 vs. 2.22 ± 0.08 kg for Cndi 13.6, p < 0.0001, a reduction of 37%. Predialysis systolic blood pressure was not influenced by Cndi, but predialysis diastolic BP was significantly lower in dialysis sessions using Cndi 13.0 mS/cm compared with 13.6 mS/cm (75.7 ± 2.1 mm Hg vs. 82.7 ± 1.8 mm Hg, p < 0.05). Postdialysis BP was significantly lower in dialysis sessions using Cndi 13.4 mS/cm compared with Cndi 13.6 mS/cm (118.9 ± 3.4 mm Hg vs. 130.1 ± 3.4 mm Hg, p < 0.05 for systolic BP; 67.8 ± 1.7 mm Hg vs. 73.8 ± 1.6 mm Hg, p < 0.05 diastolic BP). There was no significant difference in postdialysis BP between sessions using alternative Cndi. There was also no significant difference in any of the markers of hemodynamic stability between sessions using different Cndi, once established on that level.

Figure 3.
Figure 3.:
Interdialytic weight gain for all dialysis sessions using Cndi 13.0, 13.2, 13.4, and 13.6 mS/cm. Lower levels of Cndi were associated with lower weight gains. Cndi, dialysate conductivity.

Sodium balance results.

Once steady state had been achieved, IMBtotal was unchanged in dialysis sessions using different conductivities (367 ± 26 mmol vs. 414 ± 17 mmol for Cndi of 13.0 vs. 13.6, p = ns). However, IMBDiff progressively increased as Cndi decreased (Figure 4). Thus the relative contribution of diffusive and convective sodium loss to IMBtotal changed, from IMBDiff making up 21.2 ± 2% of IMBtotal when dialyzing at 13.6 mS/cm to 48.3 ± 3 % when dialyzing at 13.0 mS/cm. IMBDiff showed a strong correlation with initial Cp at each different Cndi (Figure 5).

Figure 4.
Figure 4.:
Diffusive IMB at varying Cndi, indicating significantly increased diffusive loss of sodium at lower Cndi. IMB, ionic mass balance; Cndi, dialysate conductivity.
Figure 5.
Figure 5.:
Plot of initial Cp vs. Diffusive IMB. This shows a strong correlation between these variables at each level of Cndi.

These results indicate that, at steady state, there is no difference in the total mass of sodium removed by dialysate of varying conductivity. Clearly, this will not be true for the initial dialysis session when Cndi is first changed. We analyzed the sessions in which Cndi was first reduced to 13.4 mS/cm and compared them to sessions where Cndi had been established for more than 1 week. The results demonstrate that more sodium is removed during the first dialysis session at a lower conductivity, with the increase being caused by increased diffusive loss (Diffusive IMB 98 ± 10 mmol at 13.6 mS/cm, 113 ± 8 mmol at 13.4 mS/cm, but 159 ± 23 mmol for the first dialysis session to drop Cndi to 13.4 mS/cm, p < 0.05 for the difference between first and established sessions at 13.4 mS/cm). The interdialytic weight gain and initial Cp were not significantly different when comparing the first dialysis at 13.4 mS/cm with dialysis at an established Cndi 13.6 mS/cm.

There was a trend towards lower scores in visual analog scoring of thirst severity at lowest achieved Cndi compared with baseline, but this did not achieve statistical significance (6.12 ± 0.39 vs. 4.89 ± 0.47, p = 0.08). To ensure that we had not simply selected a group of patients who had low Cp and were thus better able to tolerate low Cndi, we analyzed results grouped according to lowest achieved Cndi. This, therefore, compares results for the same patients at different Cndi levels, first at the baseline level of 13.6 mS/cm, then at the lowest Cndi achieved for that patient. Table 2 shows results for these groups, showing a similar pattern to those for dialysis sessions grouped together. As before, the reductions in final Cp and weight loss, as well as the increase in IMBDiff, were all significant. This remained true for each group of patients. The results confirm the analysis of dialysis sessions grouped together.

Table 2
Table 2:
Results table comparing baseline measurements at Cndi 13.6 mS/cm with measurements at lowest achieved Cndi.

The intervention group, analyzed as a whole, showed a similar pattern of results. Mean achieved Cndi was 13.23 mS/cm. Predialysis BP fell by a mean of 7/5 mm Hg (p < 0.05 for systolic blood pressure and diastolic blood pressure). Postdialysis BP fell by a mean of 11/4 mm Hg (p < 0.0001 for SBP, p = 0.053 for DBP). Initial plasma Cn fell by 0.21 and final plasma Cn by 0.31 (both p < 0.0001). Overall IMB did not change, whereas the diffusive IMB increased by 66 mmol (p < 0.0001)

Control Group

Similarly, to ensure that results were not caused by a change in medical, nursing, or patient behavior during the study period, we also analyzed the data for the contemporaneous control group. The results for the control group are displayed in Table 2, demonstrating no significant difference between results at the start and those obtained at the end of the trial period.


The results from the validation phase of the study confirm that plasma [Na+] is directly related to Cp, and that IMB is directly related to sodium mass removed, as estimated by plasma [Na+]. IMB as an estimate of total sodium removal has previously been validated against direct dialysis quantification, using total waste dialysate collection (r = 0.94),16 although this study used conductivity measurements of the waste dialysate. Our validation, which relied upon the measurement of plasma [Na+], also showed a close correlation between the two methods for estimating sodium flux. Assuming that measurement of plasma sodium has an accuracy of approximately 2 mmol/L and total body water an error of 10% would lead to an approximate 20–50% total error, depending upon ultrafiltration volume.23 These two factors are, therefore, likely to account for much of the variation seen in Figure 1. Despite these known errors, the two methods used showed good agreement. The known errors would also account for the difference between the agreement demonstrated by Moret (r = 0.94) and that shown by our data (r = 0.81).

An alternative explanation for the difference between sodium removal calculated from plasma sodium measurements and IMB derived from conductivity measurements would be that rapid changes in pH, chloride, bicarbonate, and potassium across a dialysis session might have a small effect upon conductivity measurements. This study suggests that the magnitude of any such effect is likely to be small, given the small relative difference between results using the two methods.

This study confirms that reduction of Cndi on an individual basis is both possible and beneficial and allows reduced interdialytic weight gains and improved blood pressure control. The benefits were seen in all groups of patients, but the most benefit was seen in those patients who achieved the lowest Cndi. All patients enrolled in the study achieved a Cndi at least 0.2 mS/cm (equivalent to 2 mmol/L of Na+) lower than the baseline of 13.6 mS/cm. Of the patients, 6 out of 16 were hemodynamically stable while dialyzing at a considerably lower Cndi of 13.0 mS/cm. Prescription of individualized Cndi, as originally suggested by Gotch,14 is therefore possible.

As in other studies,24 reduced Cndi led to a reduction in interdialytic weight gain. This was both statistically and clinically significant and was most marked in the group achieving the lowest Cndi. Thirst scores were nonsignificantly reduced with lower Cndi. The overriding stimulus for thirst is effective plasma osmolality (of which serum [Na+] is the major determinant); 25 thus the trend towards lower thirst scores is presumably related to the lower achieved Cp. Other groups have found statistically significant reduced thirst associated with lower Cndi,24 and reduced thirst would explain the observed reduction in weight gains. At least one study has found an association between very low interdialytic weight gains and worse mortality,26 which may reflect very poor dietary and fluid intake in a group with significant comorbidity. That study,26 as well as others,27 suggested that excessive interdialytic weight gains are also associated with higher mortality, possibly because of left ventricular hypertrophy (LVH), which itself is associated with chronic salt and water overload. Reducing weight gains by lowering Cndi may, therefore, be beneficial by reducing the propensity to LVH. A decreased ultrafiltration rate caused by lower weight gains may also enhance cardiovascular stability on dialysis.

The reduction in blood pressure with lower Cndi in our study is consistent with some published studies,28–30 but not others.31 De Paula et al.16 found a reduction in BP only in those with previously uncontrolled hypertension, and this may be one reason for the discrepant results between these studies. The relatively small effect seen in our study may have been caused by the lag phenomenon of several weeks between reduction of extracellular volume and reduction in blood pressure. The small numbers involved also increased the chances of a Type II error. An extended period of observation may have shown a further reduction in blood pressure. It should be emphasized that these results were achieved in a hemodynamically stable group of patients, of whom only 13% were diabetic, and 66% did not require antihypertensive medication before the study. The results are not, therefore, applicable to the entire dialysis population.

Once patients had dialyzed at the new Cndi for at least 1 week, the alterations in diffusive and convective loss balanced out, and IMBtotal was no different from baseline. In comparison, during the dialysis session when Cndi was first reduced from 13.6 mS/cm to 13.4 mS/cm, IMBtotal did increase significantly.

The IMB results suggest that when Cndi is initially reduced, more sodium is removed by diffusion, whereas convective loss is unchanged. IMBtotal, therefore, must increase, as seen in our results for an initial change in Cndi. Subsequently, although interdialytic weight gain reduces, there is an increased gradient between Cp and Cndi. The reduction in IMBconv is thus matched by an increase in IMBDiff, and hence IMBtotal reduces back to baseline steady state levels. Clearly, if an increased IMBtotal were to be maintained in the longer term, as suggested by previous studies,16 either an increase in dietary sodium intake or a relentless reduction in total body sodium would have to occur. Dietary sodium intake is unlikely to alter, and it is more likely that IMBtotal is maintained at a stable level. The patient achieves a new steady state, with lower total body sodium.

We have also demonstrated the relationship between initial Cp and IMBDiff (Figure 5). This shows a direct correlation between initial Cp and IMBDIff consistent for each different Cndi The y intercept of 0.35 mS/cm above the relevant Cndi indicates that diffusive movement of sodium can be 0 despite a small apparent gradient between plasma and dialysate. This small gradient presumably is a result of differences in effective plasma [Na+] and the Gibbs-Donnan effect.

Online conductivity monitoring is a valid, practical, and useful tool with which to study the pattern of sodium balance in patients on hemodialysis. A small reduction of Cndi was possible to some degree in all of the patients studied and appeared to be beneficial. Radical reduction in Cndi was possible in the short term in a minority of patients. Conductivity monitoring generates insights into the changes in sodium flux and plasma sodium that occur when Cndi is altered and may be helpful in monitoring the outcome of any such alteration.


Dr. C W McIntyre received an unrestricted research grant from Hospal in 2003.


1. Skroeder NR, Jacobsen SH, Lins LE, Kjellstrand CM: Acute symptoms during and between haemodialysis: The relative role of speed, duration and biocompatibility of dialysis. Artif Organs 18: 880–887, 1994.
2. Raja R, Kramer M, Barber K, Chen SL: Sequential changed in dialysate sodium during hemodialysis. Trans Am Soc Artif Inter Organs 29: 649–651, 1983.
3. Keshavaih P, Berkseth R, Ilstrup K, McMichael C, Collins A: Reduced treatment time: Haemodialysis vs. haemofiltration. Trans Am Soc Artif Inter Organs 31: 176–182, 1985.
4. Cubulsky A, Matini A, Hollomby DJ: Effects of high sodium dialysate during maintenance hemodialysis. Nephron 41: 57–61, 1985.
5. Barbour BH: Haemodialysis equipment, in Massry SG, Sellers AL (eds), Clinical Aspects of Uremia and Dialysis. Springfield, IL: Charles C. Thomas Publisher, 1976, pp. 100–145.
6. Wilkinson R, Barber SG, Robson V: Cramps, thirst and hypertension in hemodialysis patients: The influence of dialysate sodium concentration. Clin Nephrol 7: 101–105, 1977.
7. Port FK, Johnson WJ, Klass DW: Prevention of dialysis disequlibrium syndrome by the use of high sodium concentration in the dialysate. Kidney Int 3: 327–333, 1973.
8. Dominic SC, Ramachandran S, Somiah S, Mani K, Dominic SS: Quenching the thirst in dialysis patients. Nephron 73: 597–600, 1996.
9. Savage T, Fabbian F, Giles M, Tomson CRV, Raine AEG: Interdialytic weight gain and 48 hour blood pressure in haemodialysis patients. Nephrol Dial Transplant 12: 2308–2311, 1997.
10. Ozkahya M, Ok E, Cirit M: Regression of left ventricular hypertrophy in haemodialysis patients by ultrafiltration and reduced salt intake without hypertensive drugs. Nephrol Dial Transplant 13: 1489–1493, 1998.
11. Kimmel PL, Varela MP, Peterson RA, et al: Interdialytic weight gain and survival in hemodialysis patients: Effects of duration of ESRD and diabetes mellitus. Kidney Int 57: 1141–1151, 2000.
12. Bleyer AJ, Russell GB, Satko SG: Sudden and cardiac death rates in haemodialysis patients. Kidney Int 55: 1553–1559, 1999.
13. Ramdeen G, Tzamaloukas AH, Malhotra D, Leger A, Murata GH: Estimates of interdialytic sodium and water intake based on the balance principle: Differences between nondiabetic and diabetic subjects on hemodialysis. ASAIO J 44: 812–817, 1998.
14. Gotch F, Lam M, Prowitt M, Keen M: Preliminary clinical results with sodium-volume modeling of hemodialysis therapy. Proc Clin Dial Transplant Forum 10: 12–17, 1980.
15. Petitclerc T, Jacobs C. Dialysis sodium concentration: What is optimal and can it be individualized? Nephrol Dial Transplant 10: 596–9, 1995.
16. Moret K, Hassell D, Kooman JP, et al: Ionic mass balance and blood volume preservation during a high, standard, and individualized dialysate sodium concentration. Nephrol Dial Transplant 17: 1463–1469, 2002.
17. Polaschegg HD: Automatic non-invasive intradialytic clearance measurements. Int J Artif Organs 16: 185–191, 1993.
18. Bosetto A, Bene B, Petitclerc T:. Sodium management in dialysis by conductivity. Adv Ren Replace Ther 6: 243–254, 1999.
19. Petitclerc T, Bene B, Goux N, Saudon M, Jacobs C: Non-invasive monitoring of effective dialysis dose delivered to the haemodialysis patient. Nephrol Dial Transplant 10: 212–216, 1995.
20. Petitclerc T: Recent developments in conductivity monitoring of haemodialysis session. Nephrol Dial Transplant 14: 2607–2613, 1999.
21. Wong K, Xiong D, Kerr P: Kt/V in CAPD by different estimations of V. Kidney Int 48: 563–569, 1995.
22. Locatelli F, Di Filippo S, Manzoni C, et al: Monitoring sodium removal and delivered dialysis by conductivity. Int J Artif Organs 18: 716–21, 1995.
23. Mann H, Stiller S: Sodium modeling. Kidney Int 76 (Suppl.): S79–88, 2000.
24. Dominic SC, Ramachandran S, Somiah S, Mani K, Dominic SS: Quenching the thirst in dialysis patients. Nephron 73: 597–600, 1996.
25. Tomson CRV: Advising dialysis patients to restrict fluid intake without restricting sodium intake is not based on evidence and is a waste of time. Nephrol Dial Transplant 16: 1538–1542, 2001.
26. Teraoka S, Toma H, Nihei H, et al: Current status of renal replacement therapy in Japan. Am J Kidney Dis 25: 151–64, 1995.
27. Saran R, Bragg-Gresham JL, Rayner HC, et al: Nonadherence in hemodialysis: Associations with mortality, hospitalization, and practice patterns in the DOPPS. Kidney Int 64: 254–262, 2003.
28. Krautzig S, Janssen U, Koch KM, Granolleras C, Shaldon S: Dietary salt restriction and reduction of dialysate sodium to control hypertension in maintenance haemodialysis patients. Nephrol Dial Transplant 13: 552–3, 1998.
29. de Paula FM, Peixoto AJ, Pinto LV, et al: Clinical consequences of an individualized dialysate sodium prescription in hemodialysis patients. Kidney Int 66: 1232–1238, 2004.
30. Dheenan S, Henrich WL: Preventing dialysis hypotension: A comparison of usual protective maneuvers. Kidney Int 59: 1175–1181, 2001.
31. Charra B, Bergstrom J, Scribner BH: Blood pressure control in dialysis patients: Importance of the lag phenomenon. Am J Kidney Dis 32: 720–724, 1998.
Copyright © 2005 by the American Society for Artificial Internal Organs