@phdthesis{Goldau2002,
author = {Goldau, Rainer},
title = {Clinical evaluation of novel methods to determine dialysis parameters using conductivity cells},
url = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-3125},
school = {Universit{\"a}t W{\"u}rzburg},
year = {2002},
abstract = {During the last two decades an ongoing discussion about the necessary dose of dialysis brought the result that the urea based Kt/V value is significantly correlated to morbidity of the end stage renal disease (ESRD) patients. Even if it is not completely accepted, it seems to be more and more agreement of the nephrological community that for good dialysis practice Kt/V should be kept above 1.2 to 1.3 in the usual 3X4 hours per week dialysis schedule for patients without own residual clearance to assure long term quality of life, low morbidity and mortality. K is the clearance of urea the dialysis system can apply, t is the treatment time and V is the urea distribution volume of the patient, which is nearly equal to total body water. Kt/V has the unit of a drug dose (ml of drug per ml of patient volume) and therefore sometimes is called dialysis 'dose', even if this is subject of discussion because it implies that the dose can be described with only one urea related number. This work does not participate in this discussion. The premise of this work is more technical: Whatever the final result of the above discussion will be, a patient-friendly, precise cost-neutral and handy technical solution should be given to the hand of the interested nephrologist to continuously supervise the urea based Kt/V that is applied to the patient. Of course this is combined with the hope that the long term mortality can be decreased if a covering online dialysis success control is facilitated. The technical solution that has been chosen is based on the equivalence of the diffusion coefficients of sodium chloride and urea. It is central subject of the investigation if the diffusive behaviour of sodium is equal to that of urea crossing the dialysis filter membrane. The advantage that makes the principle so handy is that sodium can be measured very precise by standard conductivity cells as they are implemented in dialysis machines in large numbers. The only necessary hardware modification is a second conductivity cell downstream the dialyser to be able to measure the mass balance over the filter. This is more complicated with urea that can only be measured undergoing an enzymatic conversion to ammonium ions. The ammonium ions induce a membrane potential, which is measured with very sensitive amplifiers. A cooling chain for the enzyme must be maintained. To find and approve the conductivity based technical solution two in-vivo studies have been conducted. In the first study a conductivity step profile, varying the conductivity in static levels in a baseline - 7 min high - 7min low- baseline shape, was applied that can be utilised to measure the urea clearance very accurate. This principle has been described in 1982 in a patent application. In a sequence of 206 computer recorded dialysis sessions with 22 patients it was found that urea clearance could be electrolytically measured with a mean error+/-standard deviation of -1.46+/-4.75\% , n=494. The measurement of Kt/V according to a single pool model was of similar accuracy: 2.88+/-4.15\% . Although in accordance with other studies these findings at an average confirmed the high correlation of ionic and urea based clearance measurements, an effect was found that was not consistent with the theory that was existent so far. It was found in the first study that the accuracy of the step profile measurements were dependent of the size of the patient, in particular of the urea distribution volume. Moreover it was of relevance which part of the step profile was used: the high-low states, the baseline- low or the baseline-high states. This was a theoretical lack. Careful analysis led to the result that sodium transfer from and into the patient was the reason for the dependence. This led to the enhancement of the theory that seems to correctly describe the nature of the effect. A new demand now was to minimise the sodium transfer. This was limited using static step profiles because in the time it needs to become stable sodium is shifted. In consequence non-stable, dynamic short conductivity boli were developed that allowed to minimise the amount of sodium to be shifted to the limits of the technical resolution of the measurement systems. Also the associated mathematical tools to evaluate the boli had to be suited to the problem. After termination of this process a second study was conducted to approve the new method found. In this study with 10 patients and 93 sessions, 264 step profile measurements and 173 bolus ionic dialysance measurements it was found that the bolus measurements matched their related blood side urea clearance references with the outstanding accuracy of (error+/-SD) 0.06+/-4.76\%. The result was not significantly different (p=0.87) from the reference by student's t-test for paired data. The Kt/V reference according to the single pool variable volume urea kinetic model (sPVVUKM) was found to be matched by the bolus principle with 5.32+/-3.9\% accuracy and a correlation of 0.98. The remaining difference of 5.32\% can be attributed to the neglect of the urea generation rate. Also the step profile was found to be very precise here. The error versus sPVVUKM was 0.05\%+/-5\%, r=0.96. However it did not image the neglect of urea generation correctly. Also a two pool modelling that comprises an internal compartimentation of the fluid pools of the patient was applied to the continuously recorded data. This two pool urea kinetic model (2PUKM) is regarded to be a more precise theoretical approach and now includes the urea generation. It found the bolus principle to deviate -3.04 +/- 14.3\%, n.s., p=0.13. The high standard deviation is due to the complexity of the model. Further from the developed theory a simplified method to roughly measure the sodium distribution volume could be derived. This method was tested in-vitro versus a container with dialysate of known volume and in-vivo versus the urea distribution volume. The in-vitro results were -19.9+/-34\%, r=0.92, n.s, p=0.916. In-vivo they were found to be -7.4+/-23.2\%, r=0.71, n.s., p=0.39. Due to dilution theory the sodium and urea distribution volumes virtually appear to be very similar using this method, although they absolutely differ significantly. Facing the strong simplifications that were made before applying this theory these results seem to be very encouraging that it could be possible to develop a principle to measure not only K but also V electrolytically. This would allow a true Kt/V measurement. The empirical urea distribution volume measurement using anthropometrical formulas has been compared to analytical methods. It has been found that the use Watson's formula with a -13\% correction gives good results. The correction should be applied with great care because it increases Kt/V just on a arithmetical base to the disadvantage of the patient. Also electrolytical plasma sodium measurement was evaluated and can be measured using a mixed analytic-empirical formula with an accuracy of 4.3+/-1.2\%. In summary, conductivity based methods seem to be a convenient method to measure several dialysis parameters of some clinical interest without effort. The results of this work meanwhile are implemented with substantial numbers into commonly available dialysis machines and the experience of the first time shows that the principle is well accepted by the clinicians.},
subject = {H{\"a}modialyse},
language = {en}
}