RECOMMENDED CLINICAL PRACTICES FOR MAXIMIZING PERITONEAL DIALYSIS CLEARANCES Peter Blake,1 John M. Burkart,4 David N. Churchill,2 John Daugirdas,5 Thomas Depner,6 Richard J. Hamburger,7 Alan R. Hull,8 Stephen M. Korbet,9 John Moran,10 Karl D. Nolph,11 Dimitrios G. Oreopoulos,3 Martin Schreiber,12 and Robert Soderbloom13 University of Western Ontario,1 London; McMaster University,2 Hamilton; University of Toronto,3 Toronto, Ontario, Canada; Bowman Gray School of Medicine,4 Winston ­Salem, North Carolina; University of Illinois at Chicago,5 Chicago, Illinois; University of California,6 Sacramento, California; Indiana University,7 Indianapolis, Indiana; University of Texas Southwestern Medical Center,8 Dallas, Texas; Rush Presbyterian St. Luke's Medical Center,9 Chicago, Illinois; Baxter Healthcare Corporation,10 McGaw Park, Illinois; University of Missouri,11 Columbia, Missouri; Cleveland Clinic,12 Cleveland, Ohio; Loma Linda University School of Medicine,13 Loma Linda, California, U.S.A. reviews and original articles Data from the Canada ­U.S.A. (CANUSA) Study have recently confirmed a long-suspected linkage between total clearance and patient survival in peritoneal dialysis (PD). Recognizing that what we have historically accepted as adequate PD simply is not, the Ad Hoc Committee on Peritoneal Dialysis Adequacy met in January, 1996. This committee of invited experts was convened by Baxter Healthcare Corporation to prepare a consensus statement that provides clinical recommendations for achieving clearance guidelines for peritoneal dialysis. Through an analysis of 806 PD patients, the group concluded that adequate clearance delivered with PD can be achieved in almost all patients if the prescription is individualized according to the patient's body surface area, amount of residual renal function, and peritoneal membrane transport characteristics. Use of 2.5 L to 3.0 L fill volumes, the addition of an extra exchange, and giving automated peritoneal dialysis patients a "wet" day are all options to consider when increasing weekly creatinine clearance and KT/V. Rather than specify a single clearance or KT/V target, the recommended clinical practice is to provide the most dialysis that can be delivered to the individual patient, within the constraints of social and clinical circumstances, quality of life, life-style, and cost. The challenge to PD practitioners is to make prescription management an integral part of everyday patient management. This includes assessment of peritoneal membrane permeability, measurement of dialysis and residual renal clearance, and adjustment of the dialysis prescription when indicated. KEY WORDS: Peritoneal dialysis, adequacy, outcome. The wise general in his deliberations must consider both favorable and unfavorable factors. By taking into account the favorable factors, he makes his plan feasible; by taking into account the unfavorable, he may resolve the difficulties. - Sun Tzu, The Art of War The daily regimen of four 2-L exchanges has long been the standardized, accepted continuous ambulatory peritoneal dialysis (CAPD) prescription. Over the last few years, concern has surfaced regarding the relationship between total clearance and clinical outcomes in peritoneal dialysis (PD) patients, and several studies have confirmed that a correlation exists (1 ­3). The Canada ­U.S.A. (CANUSA) (4) prospective, multicenter cohort study has recently provided specific data that confirm the linkage between small solute clearance and patient survival. CANUSA results have led us to reevaluate our beliefs about clearance "targets," which at 50 L/wk/1.73 m²² of creatinine clearance (Ccr) and weekly KT/V > 1.7 may have been too low. A similar realization by the hemodialysis (HD) community, in 1989, resulted in an increase in the basic hemodialysis prescription in the United States (5). It is vital that PD practitioners make a similar conscious effort to increase PD prescription. In light of the CANUSA findings, the Ad Hoc Committee on Peritoneal Dialysis Adequacy met in January, 1996. This committee of invited experts was convened by Baxter Healthcare Corporation, and the objective of this committee was to provide a consensus statement that provides clinical recommendations for achieving clearance guidelines for peritoneal dialysis. It should be clearly stated that the recommendations are based on the considered opinion of the committee rather than on any validated scientific data. The aim of this committee was to evaluate clearance guidelines rather than all of the parameters, such as nutrition, that comprise PD adequacy. The above-mentioned authors are members of the Ad Hoc Committee on Peritoneal Dialysis Adequacy. Correspondence to: P. Blake, Division of Nephrology, London Health Science Centre, Victoria Campus, London, Ontario, CANADA, N6A 4G5 Received 25 May 1996; accepted 12 July 1996. BACKGROUND When one compares the clearances achieved by the "replacement" therapies (Figure 1) it is clear that these are not replacement therapies at all. Since all dialysis techniques are, at best, poor substitutes for the natural kidney, we need to strive to give as much dialysis to the individual patient as practical. In this context, interpretation of the CANUSA results is particularly enlightening. FIGURE 1 Normal kidney function in comparison to renal replacement therapy. In the CANUSA study, 680 new CAPD patients were enrolled in a prospective study in ten Canadian and four U.S. centers to determine the effect of clearance levels and nutrition on morbidity and mortality. The dialysis dose was at the discretion of the patient's individual investigator and was not increased as residual renal function (RRF) declined. In fact, mean peritoneal KT/V did not change over the two years of follow-up, and peritoneal creatinine clearance increased only slightly from 44.2 L/wk/1.73 m²² at baseline to 47.3 L/wk/1.73 m²² at 24 months (p < 0.005, baseline vs 24 months). The CANUSA study data confirm that "more is better" in peritoneal dialysis. Two-year predicted patient survival data demonstrate that lower total solute clearances are associated with higher patient mortality (Table 1). Actuarial two-year survival was 78%. In a multivariate analysis, a 5 L/week/1.73 m²² increase in total creatinine clearance was associated with a 7% decline in relative risk of death, and a 0.1-U increase in weekly KT/V was associated with a 6% decline in relative risk of death. It should be noted that changes in total creatinine clearance and KT/V in the CANUSA study are primarily due to decreases in residual renal function, rather than to alterations in peritoneal clearance. It is widely presumed, although unproven, that these findings can be extrapolated to support the existence of a similar association for alterations in peritoneal clearance. Since the CANUSA data are usually presented in tabular form, the large difference in mortality between the predicted survival for a sustained creatinine clearance of 55 and 70 L/wk/1.73 m²² has led some to consider 70 L/wk/1.73 m²² the minimum creatinine target for CAPD. However, when these data are presented graphically with additional interpolated data points, the curve is smooth and does not plateau (Figure 2). While these data support the view that more total clearance is better, the suggested target levels can be questioned for several reasons: . Greater levels of total clearance are associated with better survival, thus, selecting one level over another is somewhat arbitrary . The analysis is complicated by the fact that RRF was responsible for almost all of the variation of total clearance. . The CANUSA study was unable to demonstrate a correlation between small molecule clearance and clinical outcomes independent of RRF because there was little variation in peritoneal clearance in the study population . The data are inferences from statistical modeling (6) . The data are based on uncontrolled, nonrand-omized cross-sectional correlations . Correlations based on clearances that have been measured every six months may significantly overstate the levels at which adverse events actually occur due to wide variances in the rate of RRF decline (7) TABLE 1 CANUSA: Predicted Two-Year Patient Survival Versus Sustained Creatinine Clearance Creatinine clearance (L/week/1.73 m²²) Predicted 2-year survival (%) 95 86 80 82 70 78 55 72 40 65 FIGURE 2 The CANUSA data presented graphically, with confidence intervals that are wide and overlapping. KEY PATIENT ATTRIBUTES INFLUENCING THE DIALYSIS PRESCRIPTION To achieve the desired clearance levels, individually tailored PD prescriptions are essential. Body surface area (BSA), residual renal function, and peritoneal membrane transport characteristics are fundamental to the PD prescription. In order to confirm that it is possible to achieve the recommended clearance levels with reasonable PD prescriptions, we examined data from 806 randomly selected adult male and female PD patients from 39 U.S. centers. All 806 patients underwent a standard four-hour 2-L 2.5% dextrose peritoneal equilibration test (PET). BODY SURFACE AREA Body surface area obviously affects the amount of dialysis a patient needs. While 1.73 m²² is considered the average adult BSA, in these 806 patients (8) the median body surface area was 1.85 m²² (Table 2). In fact, 75% of the patients in this study had BSA > 1.71 m²², reinforcing the need to individualize PD prescription and utilize larger volumes of solution. RESIDUAL RENAL FUNCTION Residual renal function makes it easier to achieve clearance guidelines. For example, 1 mL/min of glomerular filtration rate (GFR), calculated as the sum of the urea and creatinine clearances divided by 2 (corrected Ccr), adds 10 L/wk of Ccr. Similarly, for each 1 mL/min of urea clearance, 0.25 is added to the total weekly KT/V for a 70-kg person. The peritoneal dialysis prescription must be adjusted as RRF declines. It is recommended that one replace overall solute clearance loss due to a decrease in RRF with a similar increase in peritoneal clearance. RRF must be measured every three months or presumed to be absent in determining the PD prescription. However, because the decline in RRF is unpredictable and may occur at different rates for different disease states, specific circumstances (e.g., antibiotic usage) may merit more frequent monitoring. PERITONEAL MEMBRANE TRANSPORT CHARACTERISTICS Peritoneal membrane transport has a significant impact on dialysis clearance and, therefore, the PD prescription. Membrane transport characteristics in the same group of 806 patients were assessed using the standard PET (9). A different distribution of peritoneal transport characteristics from that originally described by Twardowski was found in this group of patients (Figure 3), possibly because of the much higher number of patients in the current study. A small fraction (5.6%) fell into the low transport group, which is the hardest to treat with PD. Patients with high solute transport membranes comprised 10.4% of the population. Note in Table 3 that peritoneal transport characteristics were randomly distributed within each BSA group. TABLE 2 Distribution of Body Surface Area m²² (n=806) 25th percentile Median 75th percentile 1.71 1.85 2.00 FIGURE 3 Percent of patients in each PET group. SELECTING A MEASURE OF CLEARANCE The weekly Ccr and KT/V targets may not be achieved simultaneously in the individual patient for two reasons. Firstly, creatinine is secreted into the glomerular filtrate by the kidney, while urea is reabsorbed. Clearly, the residual renal creatinine clearance, even when it is estimated by the average of renal urea and creatinine clearance, is higher than the residual renal urea clearance. The relative contribution of renal creatinine clearance to total (renal + peritoneal) clearance tends to be higher than that of renal urea clearance to total KT/V. Secondly, because the transport of creatinine across the peritoneal membrane is slower than for urea, it is harder to achieve high peritoneal creatinine clearance than it is to achieve peritoneal KT/V. This is especially so in automated peritoneal dialysis (APD), when dwell times are shorter. There is no definitive proof regarding selection of the measure of clearance. Until there are data that support the use of either Ccr or KT/V exclusively as a clinical measure of PD adequacy, both should be used to assess therapy efficacy. If a discrepancy exists between these two measures, that is, one achieves KT/V but not Ccr, the patient should not be considered a failure. Instead, the targets should be considered in the total context of the patient's management through careful review and ongoing monitoring of clinical symptoms. CLINICAL ASSESSMENT AND QUALITY-OF-LIFE CONSIDERATIONS It is important to evaluate the patient frequently, taking into account clinical assessment, biochemical parameters, and life-style, because achieving adequate clearance does not guarantee decreased mortality. Clinical assessment should include evaluating the patient's uremic symptoms, nutritional state, and comorbid disease. If a patient does not meet total solute clearance targets, but clinical assessment shows the patient is thriving, consideration of life-style is important in the decision of whether or not to modify the prescription or transfer the patient to hemodialysis. Frequent ongoing assessment is required in this situation. Conversely, if a patient appears uremic despite meeting clearance targets, alterations in prescription should be considered. RECOMMENDED CLINICAL PRACTICES TO ACHIEVE CLEARANCE GUIDELINES The recommended clinical practice is to provide the most dialysis that can be delivered to the individual patient, within the constraints of social and clinical circumstances, quality of life, life-style, and cost. The recommended numeric clearance guidelines for continuous PD (treatment given 24 hours/day) are shown in Table 4. These guidelines are for total clearance, peritoneal + renal. Only when special clinical or social circumstances dictate should the first two levels of prescription ("use caution," "borderline") be acceptable. Exceptions to these guidelines should be considered when individual patient needs and quality-of-life considerations exceed the gains achieved from increasing creatinine clearance. For example, a 94-year-old patient with other systemic disease and Ccr = 50 mL/wk who is asymptomatic may choose not to be involved with a more aggressive therapy. In recommending these guidelines, it must also be noted that an increase in peritoneal clearance will be required to compensate for the decline in residual function. In situations where KT/V guidelines but not Ccr targets are met, attempts to increase Ccr should be made, but failure to do so should not be considered an indication for transfer to hemodialysis. In discussions of dialysis adequacy, it is important to point out that renal and peritoneal clearances have been considered equivalent. One milliliter/minute of corrected Ccr, or 1 U of KT/V due to residual renal function, is assumed to equal that from peritoneal clearance. These guidelines assume that an increase in peritoneal clearance will compensate for the decline in residual function, although this has not been formally proven. Most opinion leaders would agree that renal clearance is more valuable than peritoneal clearance, but this issue needs to be resolved by further research. TABLE 3 Median Values Used to Model Each BSA/PET Group Low LA HA High >2.0 m²² BSA 2.08 2.13 2.15 2.11 D/P 0.49 0.59 0.76 0.85 1.71 ­2.0 m²² BSA 1.85 1.87 1.86 1.84 D/P 0.43 0.59 0.71 0.86 <1.71 m²² BSA 1.60 1.56 1.59 1.65 D/P 0.42 0.59 0.71 0.84 BSA = body surface area; PET = peritoneal equilibration test; LA = low average transport membrane; HA = high average transport membrane; D/P = dialysate-to-plasma creatinine concentration ratio. TABLE 4 Recommended Clearance Guidelines Creatinine clearance (L/week/1.73 m²²) KT/V (weekly) Guidelines <49 <1.70 Use caution 50 ­59 1.70 ­1.89 Borderline 60 ­69 1.90 ­2.09 Acceptable >70 >2.09 Desirable ACHIEVING THE CLEARANCE GUIDELINES For some patients, achieving the recommended clearance guidelines may require a substantial increase in the dialysis prescription. This raises the concern that targets may only be achieved by an unacceptable increase in the cost of PD and/or a deterioration in the patient's quality of life. The benefits of increasing clearance must be balanced with the risks for some patients associated with larger fill volumes. The increase in intra-abdominal pressure associated with higher fill volumes may contribute to hernias for those patients prone to them and may also result in increased peritoneal fluid absorption and less net ultrafiltration. Increased glucose absorption associated with larger fill volumes is especially a consideration for patients with high transport membranes. To evaluate the options for achieving clearances, we created 12 standard "patients" by taking the median value from each of the 12 possible BSA/PET combinations (Table 3). These 12 patients were then modeled using a computer prescription program (PD ADEQUEST™™, Baxter Healthcare, Deerfield, IL) to assess which prescriptions meet the recommended guidelines and thus provide a template for an initial PD prescription for a patient of known BSA and peritoneal transport characteristics. The program, PD ADEQUEST™™, has been scientifically validated in a small well-controlled study (10,11). It has also been clinically validated for CAPD patients in a large multicenter study [see Vonesh et al. (12)]. While validation studies for APD patients are underway, results are not yet available. However, clinical validation of a kinetic model very similar to PD ADEQUEST™™ was carried out for both CAPD (n = 75) and APD (n = 25) patients by Robertson et al. (13). Their results are similar to the results reported by Vonesh and Rippe (11), with predicted clearances agreeing with clinical data to within an average difference of 10%. Although the modeling was done with PD ADEQUEST™™, there are other models available, including several commercial programs. The main point of this exercise was to demonstrate that almost all patients who develop anuria while being treated by PD can achieve acceptable targets by identifying membrane characteristics, measuring clearance, and modifying the dialysis prescription when necessary. The predicted clearances presented here do not represent absolute values but approximate levels of clearance for "typical" patients. For the majority of patients, several prescriptions were possible that met acceptable clearance, life-style, and cost criteria. It is worth noting that even patients with low transport membranes (who comprised only 5.6% of the sample studied), can be managed reasonably well with PD if they are small or have sufficient residual renal function. THE IMPACT OF DIALYSIS PRESCRIPTION ON CLEARANCE We modeled selected prescriptions for each BSA group to illustrate the relationship between therapy levels and clearance. In Figures 4 through 13 (4, 5, 6, 7, 8, 9, 10, 11, 12, 13), the modeled creatinine clearances and KT/V are shown with minimum target clearances depicted. The patients are assumed to have no residual renal function‹‹the "worst case" scenario. In each figure, three regimes are shown: standard four-exchange CAPD; three typical daytime exchanges and two exchanges during the night (one using an automated exchange device), labeled night exchange device; and APD. The following assumptions were made during modeling: . CAPD: four exchanges per day and an average ultrafiltration (UF) volume of 1.5 L . Night exchange device: three typical daytime exchanges and two five-hour exchanges overnight (one using an automated night exchange device), and average ultrafiltration volume of 1.5 L . APD: overnight treatment of ten hours with daytime dwell ("wet day": a daytime dwell infused as the last step of the APD treatment) and an average ultrafiltration volume of 1.5 L . All patients were modeled with an ultrafiltration of between 1 and 2 L/day. For the higher transport patients, the UF was closer to 1 than 1.5. UF was not universally held constant This series of figures demonstrates that you can achieve targets for patients with differing BSA and peritoneal membrane characteristics by individualizing patient prescription. Figures 4 through 7 (4, 5, 6, 7) show the benefit of replacing routine 2-L fill volumes with 2.5 L in the prescriptions of patients with BSA lower than the 25th percentile. Significantly, larger fill volumes allow almost all patients to reach clearance and KT/V targets, even when they become anuric. The main exception is the group of patients with low transport membranes. It should be noted that these "small" patients had a median BSA of 1.6 m²². A similar benefit is seen when 3-L fill volumes are used rather than 2.5 L in the patients with BSA between the 25th and 75th percentile (Figures 8 ­11: 8, 9, 10, 11). Even in the large anuric patient, a 3-L fill volume provides adequate dialysis in most individuals (Figures 12, 13: 12, 13). The same benefit should be noted for the large (>75th percentile) patient shown in Figure 12. At the end of 24 months, the CANUSA patients had a mean 14 L/wk/1.73 m²² of creatinine clearance from residual renal function (Figure 14). If this level of residual renal function is added to the creatinine clearance shown in Figure 9 for the middle-sized patient, the benefit of residual renal function can be clearly illustrated (Figure 15.) FIGURE 4 Creatinine clearance levels for patients with BSA < 1.71 m²², no residual renal function, and 2-L fill volumes. FIGURE 5 Creatinine clearance levels for patients with BSA < 1.71 m²², no residual renal function, and 2.5-L fill volumes. FIGURE 6 KT/V levels for patients with BSA < 1.71 m²², no residual renal function, and 2-L fill volumes. FIGURE 7 KT/V levels for patients with BSA < 1.71 m²², no residual renal function, and 2.5-L fill volumes. FIGURE 8 Creatinine clearance levels for patients with BSA between 1.71 and 2.0 m²², no residual renal function, and 2.5-L fill volumes. FIGURE 9 Creatinine clearance levels for patients with BSA between 1.71 and 2.0 m²², no residual renal function, and 3.0-L fill volumes. FIGURE 10 KT/V levels for patients with BSA between 1.71 and 2.0 m²², no residual renal function, and 2.5-L fill volumes. FIGURE 11 KT/V levels for patients with BSA between 1.71 and 2.0 m²², no residual renal function, and 3.0-L fill volumes. FIGURE 12 Creatinine clearance levels for patients with BSA > 2.0 m²², no residual renal function, and 3.0-L fill volumes. FIGURE 13 KT/V levels for patients with BSA > 2.0 m²², no residual renal function, and 3.0-L fill volumes. FIGURE 14 At the end of two years in the CANUSA study, residual renal function continued to supply 23% of the total creatinine clearance, at a rate of 1.4 mL/min or 14 L/week. FIGURE 15 The impact of residual renal function on creatinine clearance for a patient with BSA between 1.71 and 2.0 m²². SPECIAL CONSIDERATIONS FOR AUTOMATED PERITONEAL DIALYSIS Patients on APD, except those with high transport membranes (only 10.4% of the patients), require "wet" days. Standard APD practice should also include 2.5-L fill volumes. Increasing the overnight total solution volume overnight without an increase in time may not always provide additional clearance and may actually provide less clearance for patients with low and low average transport membranes. Thus, it is more effective to add a daytime exchange to increase clearance rather than to increase the number of exchanges delivered overnight. Figure 16 illustrates the effectiveness of different APD regimes. FIGURE 16 The effects of wet days, increasing the number of exchanges, and adding a midday exchange to APD therapy prescriptions. The difference between a dry day is illustrated clearly in the first two bars of each membrane transport type. The effect of a midday exchange is shown in the last bar. CONCLUSIONS OF THE ADEQUACY CONSENSUS GROUP Adequate clearance with PD can be achieved in almost all patients, even those with no residual renal function. The PD prescription must be individualized, taking into account the patient's body surface area, residual renal function, and peritoneal membrane transport characteristics. Assessment: To measure dialysis dose delivered, a 24-hour dialysate collection should be analyzed for creatinine and urea, to check delivered clearance at least every six months and approximately one month after a therapy change. Residual renal function should be either measured every three months or assumed to be absent to determine the appropriate prescription. It is critical to adjust the prescription to maintain adequate total clearance as residual renal function declines. To determine peritoneal membrane clearance, transport should be measured two to four weeks after initiation of PD (13) to facilitate modeling, regularly thereafter, and whenever there is an unexplained change in the 24-hour dialysate collection results, such as after severe peritonitis or suspected changes in ultrafiltration. Fill Volume: Two-and-a-half liters is the preferred fill volume for both CAPD and APD patients. For patients larger than 2.0 m²², the use of 3.0 L should be considered in the absence of residual renal function. Increasing fill volume is a more efficacious way of improving clearance than increasing the number of exchanges. This is also the easiest and most cost-effective way to increase PD clearances. CAPD: For CAPD patients with no residual renal function, a fifth exchange may be required. A convenient way to provide this is with a simple automated overnight exchange device. APD: All patients on APD, except patients with high transport membrane characteristics (10.4% of the U.S. PD population), require "wet" days. APD clearance can be raised by increasing the fill volume, lengthening the dwell time (often with a decrease in the number of cycles), or adding a daytime exchange. Utilizing a larger fill volume is more effective than increasing the number of night cycles for increasing clearance. Patients with Low Transport Membranes: Patients with low transport membranes (5.6% of the U.S. PD population) present special challenges in delivering adequate PD. These patients may be unsuitable for PD if anuric, and transfer to HD when residual renal function is absent should be strongly considered if the patient cannot meet acceptable adequacy targets. CHANGING PRESCRIPTION PRACTICES Well-prescribed PD is an excellent therapy for the majority of end-stage renal disease patients, and patients who are well dialyzed have better out comes than patients who are not. As with all medical therapies, individualizing the prescription is the key to improving the patient outcome and long-term PD therapy success. The decline in residual renal function must be balanced by a corresponding increase in peritoneal clearance achieved through a modification of the dialysis prescription. There are elements of prescription management that remain unproven today. Thus, we are relegated to utilizing information to the best of our judgment based upon the intuition of experts. Further study is required to establish which levels are required to optimize outcomes. Prescription management is not a fixed science; thus, the blending of data and clinical impression continues to be the main element in determining appropriate patient prescription. There are several research needs that are clear: (1) validation of the tools used for prescription management (an initiative which is currently under way, (2) evaluation of the impact of increasing dose in PD as residual renal function declines, (3) a better understanding of whether renal KT/V and peritoneal KT/V are equivalent, and (4) and the extent to which patient compliance has an impact on PD outcomes. The wide array of regimes available, combined with knowledgeable prescription practice, will ensure a successful patient outcome. The challenge to the individual practitioner is to make prescription management an integral part of everyday patient management. The question "What should the recommended clinical practice be?" raises a more critical question: "Who is the patient?" ACKNOWLEDGMENT Special thanks to Ed Vonesh, Teri Dunham, and Jo Tebeau for their work in modeling these patients, and to Rosalie Villano for her invaluable assistance in preparation of the manuscript. REFERENCES . Teehan BP, Schleifer CR, Brown J. Adequacy of continuous ambulatory peritoneal dialysis: morbidity and mortality in chronic peritoneal dialysis. Am J Kidney Dis 1994; 24:990 ­1001. . Brandes JC, Piering WF, Beres JA, Blumenthal SS, Fritsche C. Clinical outcome of continuous peritoneal dialysis predicted by urea and creatinine kinetics. J Am Soc Nephrol 1992; 2:1430 ­5. . Blake PG, Balaskas E, Blake R, Oreopoulos DG. Urea kinetics had limited relevance in assessing adequacy of dialysis in CAPD. In: Khanna R, Nolph KD, Prowant BF, Twardowski ZJ, Oreopoulos DG, eds. Advances in peritoneal dialysis. Toronto: Peritoneal Dialysis Bulletin Inc., 1992; 8:65 ­70. . Canada ­USA (CANUSA) Peritoneal Dialysis Study Group. Adequacy of dialysis and nutrition in continuous peritoneal dialysis: association with clinical outcomes. J Am Soc Nephrol 1996; 7:198 ­207. . Department of Health and Human Services. Highlights from the ESRD core indicators study for hemodialysis patients, January, 1996. . Blake PG. Techniques for modeling adequacy in patients on peritoneal dialysis. Am J Kidney Dis 1996; 27:750 ­3. . Gotch F, Gentile G, Keen M. The incident patient cohort study design with uncontrolled dose may result in substantial overestimination of mortality as a function of peritoneal dialysis dose. Asaio J 1996; 42:102. . Baxter Healthcare Corporation. Internal Baxter evaluation of 806 patients at 40 U.S. centers. McGaw Park, Illinois, January, 1996. . Twardowski ZJ. Clinical value of standardized equilibrium tests in CAPD patients. Blood Purf 1989; 7:95 ­108. . Vonesh EF, Lysaght MJ, Moran J, Farrell P. Kinetic modeling as a prescription aid in peritoneal dialysis. Blood Purif 1991; 9:246 ­70. . Vonesh EF, Rippe B. Net fluid absorption under membrane transport models of peritoneal dialysis. Blood Purif 1992; 10:209 ­26. . Vonesh EF, Burkart J, McMurray SD, Williams PF. PD kinetic modeling: validation in a multicenter clinical study. Perit Dial Int 1996; 16:471-81. . Robertson BC, Juhasz NM, Walker PJ, et al. A prescription model for peritoneal dialysis. ASAIO J 1995; 41:116 ­26.