Estimating equilibrated Kt/V from an intradialytic sample: effects of access and cardiopulmonary recirculations

Mass Balance Index: An Index for Adequacy of Dialysis and Nutrition

Determining adequacy of dialysis has remained a problem for the nephrologist despite the results of the National Cooperative Dialysis Study published more than 20 years ago. Urea Kinetics Modelling (UKM) which requires computer data entry is time-consuming for the dialysis staff but is the only method that has been rigorously studied. Furthermore, it is unclear today what value of Kt/V represents ideal dialysis; the technique is subject to a number of errors associated with estimation of dialyser clearance (K) and volume of distribution of urea (V) but it is useful for calculating protein catabolic rate (PCR). Methods that use urea reduction ratios (URR) is widely used because it is simpler but not always accurate and suffer from an inability to calculate PCR. Direct dialysis quantification (DDQ) can overcome a number of these problems but it is too cumbersome for routine use. Simpler methods to determine dialysateside kinetics have the advantage of solving a number of these problems and also facilitate the calculation of PCR to determine the patient's nutritional state. In our study, we have demonstrated that by taking two dialysate samples at the beginning and at the end of dialysis (2-DSM), it is possible to determine total urea removal (TUR) which is equivalent to DDQ. By taking blood samples after dialysis and before the next dialysis, it is possible to calculate the total urea generated (TUG). The ratio of TUR/TUG will provide an index of dialysis which places emphasis on removal of solute that has accumulated in the inter-dialytic interval thus re-establishing a state of equilibrium. We refer to this index as the Mass Balance Index (MBI). The MBI is also useful in helping to identify those patients whose PCR is inadequate since the mean MBI for patients with an nPCR <0.8 was 0.93 ± 0.03 vs 1.08 ± 0.02 in those with a PCR >0.8. In these two groups of patients the Kt/V was not significantly different, 1.49 ± 0.07 vs 1.53 ± 0.06, p -0.64. We suggest that the emphasis for adequacy of dialysis should shift away from Kt/V to maintaining a state of equilibrium by removing the solutes that accumulate between dialysis and by identifying those patients with an inadequate PCR.

Adaptive network based on fuzzy inference system for equilibrated urea concentration prediction

Post-dialysis urea rebound (PDUR) has been attributed mostly to redistribution of urea from different compartments, which is determined by variations in regional blood flows and transcellular urea mass transfer coefficients. PDUR occurs after 30-90min of short or standard hemodialysis (HD) sessions and after 60min in long 8-h HD sessions, which is inconvenient. This paper presents adaptive network based on fuzzy inference system (ANFIS) for predicting intradialytic (Cint) and post-dialysis urea concentrations (Cpost) in order to predict the equilibrated (Ceq) urea concentrations without any blood sampling from dialysis patients. The accuracy of the developed system was prospectively compared with other traditional methods for predicting equilibrated urea (Ceq), post dialysis urea rebound (PDUR) and equilibrated dialysis dose (eKt/V). This comparison is done based on root mean squares error (RMSE), normalized mean square error (NRMSE), and mean absolute percentage error (MAPE). The ANFIS predictor for Ceq achieved mean RMSE values of 0.3654 and 0.4920 for training and testing, respectively. The statistical analysis demonstrated that there is no statistically significant difference found between the predicted and the measured values. The percentage of MAE and RMSE for testing phase is 0.63% and 0.96%, respectively.

[Physical principles of renal replacement therapy applied to end stage renal disease patients]

"Hemodialysis" is the generic term that refers to all forms of renal replacement therapy (RRT) able to restore periodically the "internal milieu" composition in end stage renal disease patients (ESRD). RRT includes several modalities (hemodialysis, hemofiltration, hemodiafiltration) that induce basic physical principles (diffusion, convection, adsorption) via an exchange module (dialyser) and an electrolytic exchange solution (dialysis fluid). The cleansing property of the RRT depends on different factors: the treatment modality itself, the uremic toxin considered, patient's characteristic and the operational conditions (duration of treatment, session frequency, blood and dialysate flow rates). Solute instantaneous clearances reflect the dialyser's performances used in optimal conditions but not necessarily the body clearance. The effective solute body clearance is more difficult to assess in clinical practice since it includes some variables such as the treatment duration, the biological complexity of internal milieu and the variability of the patient/dialysis system interaction. The "dialysis adequacy" concept that governs the treatment efficacy in ESRD patients could not be reduced to the urea Kt/V ratio. It must integrate a selection of pertinent clinical and biological markers covering the complete spectrum of uremic abnormalities. Adequate knowledge of those basic physical principles that control the solute exchange in hemodialysis patient is highly recommended to any nephrologist who looks forward to improve treatment efficacy and reduce mortality in ESRD patients.

Resistin serum levels are increased but not correlated with insulin resistance in chronic hemodialysis patients

BACKGROUND/AIMS

Insulin resistance is a well-known phenomenon in uremia. Resistin, a recently discovered insulin inhibitor secreted by adipocytes, is associated with obesity and insulin resistance in mice. Adiponectin, also secreted by adipocytes, is known to reduce insulin resistance in humans. The aim of the present study was to address the hypothesis that changes in resistin or adiponectin serum levels may relate to body composition and to insulin resistance in patients with end-stage renal disease.

METHODS

In a cross-sectional study, 33 non-diabetic patients (24 males and 9 females, mean age 61.5+/-15.8 years) with end-stage renal disease on chronic hemodialysis (treatment duration 41+/-31 months) that lacked signs of infection were enrolled. The control group consisted of 33, matched for age, sex and body mass index (BMI), healthy volunteers (22 males, 11 females, mean age 62.6+/-12.1 years). BMI (kg/m(2)) was calculated from body weight and height. Body fat (%) was measured by means of bioelectrical impedance. Blood samples were taken always in the morning after a 12-hour fasting period before and after the hemodialysis session. Resistin and adiponectin serum concentrations were measured by enzyme immunoassays and insulin by an electrochemiluminescence immunoassay. The post-treatment values were corrected regarding the hemoconcentration. The homeostasis model assessment index (HOMA-R) was calculated as an estimate of insulin resistance from the fasting glucose and insulin serum levels.

RESULTS

Pre-treatment resistin serum levels were significantly increased in hemodialysis patients compared to healthy controls (19.2+/-6.2 vs. 3.9+/-1.8 ng/ml; p<0.001). Hemodialysis did not alter resistin levels, as pre- and post-treatment levels were not different when corrected for hemoconcentration (19.2+/-6.2 vs. 18.7+/-5.0 ng/ml; p=0.54). Adiponectin levels were also increased in hemodialysis patients compared to healthy controls (25.4+/-21.5 vs. 10.5+/-5.9 microg/ml; p<0.001). A significant inverse correlation was observed between the serum adiponectin levels before the hemodialysis session on the one hand and the BMI (r=-0.527, p=0.002), the HOMA-R (r=-0.378, p<0.05) and the fasting insulin levels (r=-0.397, p<0.05) on the other. However, no significant correlation was observed between serum resistin levels on the one hand versus HOMA-R index (3.2+/-3.9 mmol.microIU/ml; r=-0.098, p=0.59), insulin levels (13.3+/-14.4 mU/l; r=-0.073, p=0.69), glucose levels (89+/-13 mg/dl; r=-0.049, p=0.78), BMI (25.6+/-3.7 kg/m(2); r=-0.041, p=0.82) and body fat content (26.4+/-8.4%; r=-0.018, p=0.94) on the other hand.

CONCLUSION

Resistin serum levels are significantly elevated in non-diabetic patients with end-stage renal disease that are treated by hemodialysis. The hemodialysis procedure does not affect the resistin levels. Along with previous observations in patients with renal insufficiency in the pre-dialysis stage, our findings implicate an important role of the kidney in resistin elimination. However, increased resistin serum levels in hemodialysis patients are not related to reduced insulin sensitivity encountered in uremia.

Assessing the utility of the stop dialysate flow method in patients receiving haemodiafiltration

BACKGROUND

The stop dialysate flow (SDF) method of post-dialysis urea sampling is the most commonly used method in the UK. It can also be used with a published formula to predict 30 min equilibrated urea accurately. The method has not been validated in patients undergoing haemodiafiltration (HDF). Given the increased use of HDF across Europe, we felt it prudent to assess the utility of the SDF method and prediction equation in this modality.

METHODS

Fourteen patients from two renal units were studied. Blood samples were taken at 1 min intervals from the arterial side of the dialysis circuit in the first 5 min after HDF had ceased whilst blood circulation continued. A peripheral sample was taken from the contralateral arm immediately after HDF had ceased and a 30 min sample was taken from the arterial needle. These samples were used to assess the utility of 5 min arterial blood urea and the 30 min prediction formula, respectively.

RESULTS

Blood urea measured from the arterial circuit at 5 min correlated closely with the contralateral sample taken immediately post-HDF, with no significant difference (6.45+/-2.11 vs 6.52+/-2.19 mmol/l, P = 0.39). The use of 5 min arterial blood urea and prediction formula allowed an accurate prediction of 30 min urea (R2 = 0.96).

CONCLUSIONS

The use of the SDF method with a 5 min post-HDF arterial sample is valid in patients receiving HDF. The previously published prediction formula for estimating 30 min urea is also valid using the 5 min post-HDF sample.

Mathematical analysis of urea rebound in long-term hemodialysis patients

Quantifying hemodialysis (HD) treatment requires knowledge of the equilibrated concentrations of the post-HD small molecule rebounds. However, measurement of the equilibrated concentrations is only possible after resting in bed after HD for at least 30 min, and this is often impractical. Therefore, we have analyzed mathematically the time course of post-HD urea rebound, and from this, have derived a new formula for predicting its equilibrated concentration. The blood urea nitrogen (BUN) was measured at 10 time points (immediately following HD, and 0.5, 2.5, 5, 7.5, 10, 15, 20, 25, and 30 min post-HD) in 12 anuric HD patients. The absolute change in the urea rebound (DeltaeqBUN) was approximated (DeltaestBUN) using the equation: DeltaestBUN = b -[1-exp x (-c x time (min))] + a x time (min). After the good correlation between DeltaeqBUN and DeltaestBUN, we compared the value of DeltaeqBUN measured at 30 min (DeltaeqBUN(30)) with that calculated (DeltaestBUN(30)) using only four sample points (immediately following HD, and 2.5, 5 and 10 min post-HD). Based on this result, we tried to predict post-HD BUN at 30 min (estBUN(30)). This study was undertaken to determine whether estBUN(30) may be representative of the equilibrated BUN (eqBUN(30)), and to compare with Kt/V using estBUN(30) and eqBUN(30). There was a significant correlation between DeltaeqBUN and DeltaestBUN (0.97 < r < 0.99, P < 0.001). Thus, there was a significant positive linear correlation between eqBUN(30) and estBUN(30) (eqBUN(30): 25.7 +/- 2.25 mg/dL, estBUN(30): 26.3 +/- 2.31 mg/dL; r(2) = 0.99, P < 0.001). A Kt/V measurement was obtained with single pool model using BUN just after HD (Kt/V(sp)), eqBUN(30) (Kt/V(eq)), and estBUN(30) (Kt/V(est)), and with double pool model using Kt/V(sp) (Kt/V(dp)) and was compared with them. Though Kt/V(sp) was significantly higher than Kt/V(eq) (1.26 +/- 0.08 vs. 1.09 +/- 0.07, P < 0.001), there were no differences among Kt/V(eq), Kt/V(est) and Kt/V(dp) (Kt/V(est): 1.06 +/- 0.07, Kt/V(dp): 1.10 +/- 0.07) and all values were clinically acceptable. Furthermore, there was a significant positive linear correlation between Kt/V(eq) and Kt/V(est) (r(2) = 0.98, P < 0.001). In conclusion, we have devised the method to predict equilibrated BUN and calculate double pool Kt/V, which requires samples up to 10 min post-HD.

Two Spent Dialysate Samples are Sufficient for Hemodialysis Efficacy Assessment

Introduction

The measure of dialysis efficacy is expressed as Kt/V value (calculated from predialysis and postdialysis blood urea concentration). The aim of this study was to assess the possibility of direct calculation of Kt/V value from two spent dialysate samples by using the regular blood-based Kt/V calculation formula with dialysate samples used as surrogates for blood samples, and to detect the most appropriate couple of dialysate samples for Kt/V estimation.

Patients and Methods

Fifty-two single hemodialysis treatments in 34 anuric patients on chronic bicarbonate low-flux hemodialysis were observed. Kt/V values according to Daugirdas formula from two blood samples and from two dialysate samples were calculated.

Results

Kt/V values calculated according to Daugirdas 2nd generation formula from blood samples (Kt/Vsp Daugirdas) were in significant correlation with all Kt/V values obtained from two spent dialysate samples. The highest correlation coefficient (r = 0.74, p & 0.001) and the least standard error of mean of the differences were found between Kt/Vsp Daugirdas and value obtained with substitution of urea concentration from dialysate samples taken 60 minutes after dialysis start and at the end of the dialysis into Daugirdas 2nd generation formula (Kt/VDCD(60)-CD(e)), which can be expressed as a equation of linear regression y = 0.47 + 0.86x. The highest correlation coefficient (r = 0.74, p & 0.001) was found between Kt/Vsp Daugirdas values equilibrated according to Daugirdas rate formula, and Kt/VDCD(60)-CD(e) value, which can also be expressed as an equation of linear regression y = 0.43 + 0.73x.

Conclusion

The results of this study clearly show the sufficiency of only two spent dialysate samples for direct estimation of the Kt/V values, with no blood sample required.

Predicting 30-minute postdialysis blood urea concentrations using the stop dialysate flow method

The stop dialysate flow (SDF) method has been the recommended method of postdialysis urea sampling by the Scottish Renal Association since November 1998. However, this method does not lend itself to calculation of Kt/V using currently favored formulas, which require either a 30-minute postdialysis sample or a 20-second "slow flow" sample. We, therefore, derived a formula that uses a 5-minute postdialysis urea sample using the SDF method to estimate the urea concentration at 30 minutes. Blood samples were obtained from 70 hemodialysis patients immediately before dialysis and at 0, 5, and 30 minutes postdialysis. Half of the patients from each unit were randomly selected to form the linear regression equation: Estimated 30-minute urea concentration = 1.06 x (5-minute urea concentration) + 0.22. This equation was validated using the data from the remaining 35 patients. This showed a very close correlation between measured and estimated urea concentration at 30 minutes (R2 = 0.97), and a Bland-Altman plot confirmed this close relationship. The sensitivity, specificity, positive, and negative predictive values of this equation were high when used to estimate 30-minute urea reduction ratio (URR) greater than 65% (100%, 85.7%, 97%, and 100%, respectively) and 30-minute Kt/V greater than 1. 2 (96.7%, 100%, 100%, and 80%, respectively). The coupling of the SDF method with the above formula combines the advantages of simple and reproducible postdialysis blood sampling with an accurate estimation of the 30-minute postdialysis blood urea concentration, URR, and Kt/V. This method should be a useful tool for comparative audit of hemodialysis adequacy.

Prediction of equilibrated urea in children on chronic hemodialysis

Urea rebound (UR) after hemodialysis (HD) requires the use of equilibrated urea (Ceq) instead of immediate end-dialysis urea (Ct) for correct quantification of HD, which is impractical. A new formula for predicting Ceq in children is suggested in our study. Thirty eight standard pediatric HD sessions (single pool Kt/V = 1.70 +/- 0.35, K = 4.65 +/- 1.14 ml/min/kg, UF coeff. = 3.2-6.2 ml/h/mm Hg, t = 3.80 +/- 0.46 h) in 15 children (M: 6, F: 9), ages 14.5 +/- 3.28 years were analyzed. Blood samples were taken: before, 70 min from the start, at the end, and 60 min after the end of HD sessions. After correlating UR (20.32 +/- 7.74%) to various HD parameters, we found that it was mainly determined by HD efficiency parameters. Therefore we correlated Ceq to HD efficiency parameters (Ct, urea reduction ratio, Kt/V, and K/V) and found a very high correlation between Ct and Ceq (r = 0.973). Linear regression analysis was used to further investigate this relationship, and a new formula to predict Ceq from Ct was obtained (Ceq = 1.085 Ct + 0.729, R2 = 0.946, SE = 0.49, absolute residuals = 0.38 +/- 0.29 mmol/L). In a validation study (10 HD sessions with new set of urea blood samples) the results obtained by the new formula were compared with measured values of Ceq and those obtained by the Smye formulae. Values predicted by the new formula (9.91 +/- 2.92 mmol/L) were not significantly different from the measured values (10.33 +/- 3.44 mmol/L). Absolute error of the new formula was 0.78 +/- 0.73 mmol/L, median 0.65; ie., 6.93 +/- 5.3%, median 7.7%. Ceq predicted by the Smye formulae (10.95 +/- 4.18 mmol/L) also did not significantly differ from the measured values, but absolute error of predicted values was markedly higher (1.21 +/- 0.90 mmol/L, median 0.89; 11.73 +/- 7.72%, median 10.11%; p < 0.05). When predicted Ceq was used for calculating equilibrated Kt/V (eKt/V), the new formula resulted in lower absolute error (0.09 +/- 0.07, median 0.08) than the Smye method (0.14 +/- 0.08, median 0.12). We conclude that our simple formula is sufficiently accurate in predicting Ceq in standard pediatric HD and that it is more accurate than the existing Smye formulae, while requiring only pre- and post-HD urea samples. We suggest the use of the new formula for predicting Ceq, which can then be used instead of Ct for a more accurate estimation of double pool Kt/V, URR, V, and PCR.

Improved clearance of iohexol with longer haemodialysis despite similar Kt/V for urea

BACKGROUND

The efforts to improve the quality of haemodialysis (HD) has renewed the interest in the consequences of blood-flow distribution for removal of solutes.

METHODS

To test the effects of HD time per se, 10 patients were studied in a cross-over fashion with HD for 3 h and 1 week later for 6 h, with similar blood urea Kt/Vs, achieved by adjusting the blood flow rate to 290 and 120 ml/min respectively. Injections of iohexol (MW 821 Dalton) were given 2 days prior to the dialysis sessions. Blood samples were taken before, during (6/HD), 1 and 24 h after the HD and analysed for concentrations of urea and iohexol. A urea on-line monitor (Gambro) was used for continuous recordings and sampling of dialysate.

RESULTS

According to the study design the blood Kt/V for urea (Daugirdas II) was similar for 3 and 6 h HD, close to 1.0 (n.s), while the removed mass of urea showed that Kt/V was slightly and significantly higher for the 6 h HD. The 'apparent' mass of iohexol, defined as plasma concentration times estimated distribution volume, fell to 29% and 21% of pre-dialysis levels after 3 h and 6 h HD, respectively (P<0.01), but increased after HD, and more so after the short dialysis, reaching 46% of the predialysis mass 24 h after 3 h HD vs. 36% after 6 h HD (P<0.05). The removed mass of iohexol was 920+/-110 mg with 6h HD and 700+/-81 mg with 3h HD, (P<0.01). Thus, the longer dialysis removed 32% more iohexol despite similar blood Kt/V for urea.

CONCLUSION

The treatment time per se affects solute removal despite similar blood Kt/V for urea. This is particularly true for an intermediate-size molecule like iohexol.

Haemodynamic Alteration in Patients Undergoing Chronic Haemodialysis

Fifteen elderly patients, 13 of them undergoing chronic haemodialysis, 1 acute and 1 coming from Continuous Ambulatory Peritoneal Dialysis (CAPD) either with no significant cardiovascular alteration or presenting various cardiovascular pathologies were studied to investigate the possibility of onset of hypotensive episodes during dialytic treatment depending on cardiac or vascular alteration in the patients. Monitoring of the arterial pressure on the contralateral arm and on the lower limbs by using the Takeda System, made it possible to compute the Windsor Index (WI). The figures obtained were correlated to the Ejection Fraction Index (EFI) to investigate the relation between WI alteration and haemodynamic variations in the patient. The results show that cardiothoracic recirculation is much more present in those patients with pathologies that affect EFI which worsens during dialysis due to the loss of fluid. Moreover the results obtained from the two patients with temporary access and no evident cardiovascular pathology show the constancy of the haemodynamic parameters throughout the dialytic treatment.

Quasi-steadiness approximation for the two-compartment solute kinetic model

We analytically solved the equation of the variable volume, two-compartment solute kinetic model (TCSKM). From the solution, we constructed an expression of weekly concentration profiles developing in the patient's body by routine hemodialyses. Obtained formulas can be used to calculate Kt/V, solute reduction index (SRI), the solute generation rate (G) per unit distribution volume (V), and a mass transfer coefficient (MTC) between the two compartments. To estimate these parameters, the formulas only need three-point data during a dialysis, that is, pre-, one-hour, and post-dialysis solute concentrations instead of four that would otherwise be needed. A 48 hour data point is not required. The weekly concentration profiles can be easily calculated by the formulas. As examples of clinical applications, we calculated Kt/V, G/V, and SRI of urea, Cr, and uric acid using plasma data of 121 hemodialyzed patients. Then the results were compared with the single-compartment solute kinetic model (SCSKM). The obtained mean MTC/V values, that is, 1.08 (1/hr) for urea, 0.53 (1/hr) for Cr, and 1.11 (1/hr) for uric acid, were consistent with the previous works. SCSKM overestimated the mean G/V by 7.1%, 15.9%, and 10.0%, and the mean SRI by 6.7%, 18.6%, and 10.0%, for urea, Cr, and uric acid, respectively. The solute distribution volume ratio of TCSKM to SCSKM, (V)TCSKM/(V)SCSKM, depended on the value of MTC/V and the hemodialysis duration. Using pedometers, we measured the total number of steps the patients took during a week. We found that the total number of steps in a week was significantly correlated with the Cr generation rate (r = 0.285, P < 0.03), but that it was not significantly correlated with the other generation rates (r = 0.204, P > 0.09 for urea, and r = 0.209, P > 0.08 for uric acid). These data suggest that the Cr generation rate is related to the patient's physical activity. We conclude that the formulas can estimate an adequate dialysis prescription for the hemodialyzed patient.

Use of the medical differential diagnosis to achieve optimal end-stage renal disease outcomes

Compared with the general population, end-stage renal disease (ESRD) patients continue to have a higher than expected morbidity and mortality. Hypoalbuminemia, anemia, hypertension, and inadequate dialysis are all thought to contribute to the high morbidity and mortality among ESRD patients. Anemia algorithms should help to standardize the approach to anemia and the use of recombinant human erythropoietin (rHuEPO), but clinicians still must review each patient individually, searching for and treating the multitude of interrelated factors that affect rHuEPO responsiveness. Hypoalbuminemia is a very strong predictor of increased morbidity and mortality in dialysis and nondialysis patients. The causes of hypoalbuminemia are multifactorial, and diagnosis of the cause of hypoalbuminemia is usually elusive. The basis of the poorer survival in US dialysis patients remains controversial, but inadequate dialysis has been implicated. To assure adequate dialysis, the dialysis prescription must be individualized for each patient, and delivered dialysis must be routinely monitored. Hypertension is associated with left ventricular hypertrophy, which is also an important determinant of survival in ESRD patients. Hypertension should be treated in ESRD patients in conjunction with other interventions that are known to reverse left ventricular hypertrophy. Special efforts must be made in the medical management of hypoalbuminemia, anemia, hypertension, and dialysis treatment adequacy to improve survival in patients with ESRD.

Screening for extreme postdialysis urea rebound using the Smye method: patients with access recirculation identified when a slow flow method is not used to draw the postdialysis blood

To look for patients with extreme urea rebound, we drew intradialytic samples one third of the way into dialysis during routine modeling for 3 months. The samples taken postdialysis were obtained after stopping the blood pump, without any slow flow period. Using the Smye equations, the intradialytic urea level was used to predict urea rebound, expressed as Kt/V-equilibrated minus Kt/V-single pool (deltaKt/V). Results were averaged for the 3-month period in 369 patients. Mean estimated deltaKt/V was -0.20 +/- 0.13, which was similar to but slightly higher than the predicted value (-0.6 x K/V + 0.03) of -0.19 +/- 0.04. In 27 patients, extreme rebound (mean deltaKt/V < -0.40) was found. Sixteen of these patients consented to further study, but only after access revision in four patients. In these patients, additional slow flow samples after 15 seconds and 2 minutes of slow flow, respectively, were drawn one third of the way into dialysis and postdialysis, and a sample was drawn 30 minutes after dialysis. On restudy, postdialysis rebound was still high with full flow samples deltaKt/V = -0.40 +/- 25, but was much lower (-0.18 +/- 0.07) and similar to predicted rebound (-0.19 +/- 0.05; P = NS) when based on 15-second slow flow samples. Eight of the 16 had marked (>15%) access recirculation by urea sampling, and deltaKt/V based on full flow post samples correlated with access recirculation (r = -0.91). The results suggest that the Smye method is valuable for identifying patients with aberrantly large postdialysis rebound values. When the postdialysis samples are drawn without an antecedent slow flow period, most patients with extreme rebound values turn out to have marked access recirculation.