 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
CLINICAL SCIENCE |
*Charing Cross Hospital, London, United Kingdom; North Staffordshire Hospital, Stoke on Trent, United Kingdom; Maelor General Hospital, Clydd, Wales, United Kingdom; Centre Hospitalier Louise Michel Evry, Evry, France; ||Hospital Amau de Vilanova, Lerida, Spain; ¶Klinikum Darmstadt, Darmstadt, Germany; #Baxter Renal Division Europe, Brussels, Belgium; and **Baxter Healthcare Corporation, Round Lake, Illinois
Correspondence to Dr. Edwina A. Brown, Faculty of Medicine, Imperial College London, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK. Phone: 44-20-8846-7590; Fax: 44-20-8846-7589;
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
60 L/wk per 1.73 m2 and ultrafiltration (UF) 750 ml/24 h during the first 6 mo. Baseline solute transport status (D/P) was determined by peritoneal equilibration test. At 1 yr, 78% and 74% achieved Ccrea and UF targets, respectively; median drained dialysate volume was 16.2 L/24 h with 50% of patients using icodextrin. Baseline D/P was not related to UF achieved at 1 yr. At 2 yr, patient survival was 78% and technique survival was 62%. Baseline predictors of poor survival were age (>65 yr; P = 0.006), nutritional status (Subjective Global Assessment grade C; P = 0.009), diabetic status (P = 0.008), and UF (<750 ml/24 h; P = 0.047). Time-averaged analyses showed that age, Subjective Global Assessment grade C and diabetic status predicted patient survival with UF the next most significant variable (risk ratio, 0.5/L per d; P = 0.097). Baseline Ccrea, time-averaged Ccrea, and baseline D/P had no effect on patient or technique survival. This study shows that anuric patients can successfully use APD. Baseline UF, not Ccrea or membrane permeability, is associated with patient survival. E-mail: e.a.brown@imperial.ac.uk |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
Reanalysis of data from the CANUSA study suggests that peritoneal and renal clearances are not equivalent (6). The current guidelines are therefore speculative for anuric patients on PD. Indeed, anuric patients are a difficult group to dialyze effectively using CAPD (7). Not only may membrane permeability increase with time on PD with a consequent decrease in achieved UF (8), but also it is difficult to increase dialysis small solute clearance using CAPD because of the practical difficulties in increasing exchange volume and/or number. Automated peritoneal dialysis (APD) is a possible alternative for dialyzing anuric patients. The nighttime component of APD enables more frequent exchanges with reduced cycle length, thereby enhancing UF. Solute clearance can be increased by using higher volume exchanges and by adding daytime exchanges (9), and UF can be increased by the use of icodextrin for the long-day dwell (10). Even so, there is still doubt as to whether adequate clearances and successful clinical outcomes can be achieved, particularly in patients with large body surface area (11).
There are some data about the outcomes of anuric patients on PD, mostly CAPD, as very few patients in any of the studies were on APD. The largest prospective study of PD in anuric patients is from Hong Kong (12); all patients were on CAPD, and mean creatinine clearance (Ccrea) was low at 43.7 L/wk per 1.73 m2. Two-year patient and technique survival in this study was 68.8% and 61.4%, respectively. Anuric patients were included in another study from Hong Kong with a 2-yr patient survival of 80% (13). Both of these studies suggest that outcome may be related to baseline Kt/V, as does retrospective data from Toronto (14). Another retrospective study suggested that Kt/V may be related to hospitalization rates but not mortality in 142 anuric patients (15). In contrast, the relationship between small solute clearance and survival was not confirmed by the ADEMEX study (16), a large, randomized, prospective study with 965 patients, more than half of whom were anuric at baseline.
The European APD Outcome Study (EAPOS) is a 2-yr prospective, observational, multicenter study of anuric patients receiving APD. The aim of the study was to determine the factors that affect patient and technique survival and thereby to arrive at guidelines for the treatment of such patients. Unlike previous studies, dialysis prescription was altered throughout the study, aiming to achieve both small solute clearance (Ccrea >60 L/wk per 1.73 m2) and UF (>750 ml/24 h) targets. The baseline characteristics of the patients recruited into this study have been reported elsewhere (17).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
Study Design
The aim of the study was to determine the impact on technique and patient survival of achieving the predetermined targets of a total Ccrea of 60 L/wk per 1.73 m2 and a UF rate of 750 ml/24 h. Before initiation of the study, the investigator group established guidelines for the starting APD prescription for anuric patients (Table 1). These guidelines were based on a combination of personal clinical experience of the investigators, modeling using PD Adequest, and a considerable amount of discussion. Individual centers determined the appropriate APD prescription to achieve these targets. At the baseline assessment and at 2 monthly intervals thereafter, patients who did not achieve either or both of these targets were identified at each center by the central data collection office. The dialysis prescription of each of these patients was then examined and, when possible, modified to attempt to improve Ccrea (by increasing the volume of fluid cycled overnight or by adding in daytime exchanges) and/or UF (by increasing dextrose concentration of one or more exchanges or using icodextrin for the long-day dwell). The aim was to optimize the APD prescription over the first 6 mo of the study. In addition, targets were set for BP and biochemical and hematologic parameters as defined by the UK Renal Association (18). Recruitment to EAPOS was completed in April 2000, and the final patient completed the study in April 2002.
|
Peritoneal Dialysis Measurements
Solute Clearance.
Weekly Ccrea was calculated from creatinine concentration in plasma, urine, and dialysate; 24-h urine volume, and 24-h dialysate volume using PD-Adequest (Baxter Healthcare Corporation, Brussels, Belgium). Residual renal function was measured as mean of urea and Ccrea.
UF Rate.
UF rate was calculated as the difference between the volume of total dialysate infused (including both nighttime and daytime fluid) and the volume drained over 24 h for the same period as the solute collection.
PET.
The PET was performed as described previously, and the recognized classification (20) was used to subdivide patients on the basis of the results.
Comorbidity.
The number of comorbidities was counted for each patient to calculate the Stoke index of comorbidity (21). This index was recently validated by an analysis of outcomes from the NECOSAD study (22).
Data Management
Peritoneal, urine, and dialysate concentrations of creatinine and all other blood tests were measured in each local laboratory. Results of these and all other demographic data were collated in the central EAPOS office (Baxter Healthcare Corporation) where the PET results and Ccrea were calculated. Individual centers were notified of patients who did not achieve the Ccrea and/or UF targets at a visit so that adjustments could be made in the APD regimen.
Statistical Analysis
Baseline and Longitudinal Data.
Parametric data are presented as mean values with their SD, and nonparametric data are presented as median values with their range. Between-group comparisons were performed using the Mann-Whitney U test, and longitudinal changes in variables were analyzed using paired and unpaired ANOVA.
Univariate Survival Analysis.
For comparing survivals according to baseline characteristics, patient, combined patient/technique, and pure technique survival were calculated using life table analysis using log rank statistics. Patients were censored at the point of transplantation or if lost to follow-up. An important feature of this study is that all patients were followed for the full 24 mo unless they reached an end point (death, technique failure, or transplantation). For calculating pure technique survival, patients were censored at death and transplantation.
Multivariate Survival Analysis.
Cox proportional hazards regression was used in three stages. First, baseline characteristics were added sequentially to a model that included age, body surface area, gender, comorbidity grade (categories 0, 1 to 2, and >2) and SGA (treated as three categories, A, B, and C). Second, as this was a study with planned interventions for UF and Ccrea, these variables, including total and peritoneal Ccrea, were treated as time-dependent covariates. As reanalysis of the CANUSA study (6) showed benefit from even small differences in urine volume, residual Ccrea was included as a separate time-dependent covariate. Finally, a forward step-wise Cox regression model was constructed in which only covariates with a P < 0.5 are included in their order of importance according to their likelihood score (
2) (23,24).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
Baseline demographicsof patients enrolled in the study are shown in Table 2. The duration of total dialysis time before recruitment into EAPOS was long (median, 37.8 mo). Of note, 64 (36%) of the patients had previously received hemodialysis for some period of their dialysis history. Fifteen percent of patients had diabetes, and 42% had cardiovascular disease at baseline. Mean D/P creatinine on PET at baseline was 0.74 ± 0.12; the distribution of patients according to transport status is shown in Table 1, with no patients being in the low transport group. Only seven patients were severely malnourished at baseline as defined by SGA grade C.
|
Details of dialysis volume, Ccrea, and UF over the course of the study are given in Table 3. At baseline, median total was 63.4 L/wk per 1.73 m2, UF rate was 1.09 L/d, drained dialysate volume was 15.7 L/d, and 45% of the patients were using icodextrin during their daytime dwell. For the 64 patients with Ccrea below target at baseline, a significant increase in prescribed volume (14.2 L/d to 15.1 L/d; P < 0.001) resulted in a mean increase in Ccrea from 49.7 to 57.5 L/wk per 1.73 m2 by 6 mo (P < 0.001). In contrast, patients with Ccrea above target at baseline had no change in dialysis dose, and Ccrea was unchanged during the first 6 mo (70.9 versus 70.1 L/1.73 m2; P = 0.34). The increase in dialysis dose below target is reflected by the increase in dialysate volume at 6 mo (Table 3). This is also reflected in the higher percentage of patients reaching the target, as shown in Table 4.
|
|
|
Other Targets
Quality of dialysis management (BP, hemoglobin, calcium, and phosphate control) was measured against the 1997 UK Renal Association standards (16). The results achieved are shown in Table 4. A total of 77% of patients were treated with human recombinant erythropoietin therapy.
Peritonitis Rate
The overall peritonitis rate during the study was 139 episodes in 2621 patient-months, or one episode in 18.8 mo. The causes of peritonitis were coagulase-negative staphylococci (40 patients [29%]), Gram-negative organisms (34 [24.5%]), Staphylococcus aureus (12 [9%]), Streptococcus viridans group (12 [9%]), fungal (4 [3%]), culture negative (21 [15%]), and unknown (16 [11%ü.
Causes of Dropout
A total of 120 patients dropped out of the study (Table 5). Twenty-four (20.3%) of these patients received a transplant; 31 (26.3%) died, mainly of cardiovascular disease. Infection was the second most common cause of death, with four patients dying of peritonitis. Transfer to hemodialysis occurred in 51 (43.2%) patients; in 22 patients, the cause was peritonitis. Two of the patients who transferred because of peritonitis died within 1 mo of transfer. Other reasons for dropout were partial recovery of residual renal function and loss to follow-up.
|
|
|
|
|
. Time-averaged UF was used in both models; however, the last value ateach 6 mo for residual and peritoneal Ccrea was used in the model for Table 8 and the time-averaged total Ccrea was used in the model for Table 9. In each case, age, an SGA grade C, and diabetic status predicted outcome, with UF the next most significant variable.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
The 2-yr patient survival of 78% and technique survival of 62% in this study are identical to those reported from the NECOSAD study (hemodialysis and PD) (22). Patients in EAPOS were representative of the European general dialysis population, with a similar age range and prevalence of diabetes and cardiovascular comorbidity to the NECOSAD study (22). A possible criticism of EAPOS is that by recruiting from a predominantly prevalent patient cohort, the patients who were selected were survivors. In fact, their survival was similar to that of the incident PD patients of the NECOSAD study (25), which along with other studies (26) have shown that survival of incident and prevalent patients are identical.
The results from EAPOS differ from those of the only other prospective study of anuric patients on PD (12), in which outcome was determined by small solute clearance. Mean Ccrea in that study was only 43.7 L/wk per 1.73 m2. Patient survival (68.8% at 2 yr) was lower than that of EAPOS but was similar to that of the ADEMEX study in which approximately half of the patients were anuric (16). ADEMEX was a large, prospective, controlled trial of 965 patients on CAPD who were randomized to two different target levels of peritoneal Ccrea. Achieved peritoneal Ccrea were different in the two groups (33rd to 67th percentile values, 42.5 to 49.1 L/wk per 1.73 m2 and 53.4 to 60.5 L/wk per 1.73 m2); however, 2-yr patient survival was similar in both groups (68.3 and 69.3%, respectively), again suggesting that survival is independent of small solute clearance. These results, though, are in contrast to other prospective and retrospective studies (3,27,28) that show increasing mortality with declining total Ccrea. In each of these studies, the decline in total Ccrea was related to loss of residual renal function as no change was made to the dialysis prescription.
Although it is difficult to compare results from different studies, it could be suggested that there is a minimum level of small solute clearance that has to be achieved to avoid morbidity directly related to uremia. Indeed, in the ADEMEX study, uremia-related deaths did occur in the low clearance group. However, once this threshold has been reached, a further increase in small solute clearance has no survival advantage, at least in the range achievable with current APD techniques. The results from EAPOS will certainly add to the debate about the need to focus on small solute clearance in PD management guidelines (29–33).
EAPOS was designed to be a real-life observational study. The only exclusion criterion was a life expectancy of <6 mo. There was no age limit, and patients who would not change their dialysis regimen to meet targets remained in the study. To be included in the study, a patient’s residual renal function had to be <1 ml/min per 1.73 m2; therefore, although some patients had some urine output at baseline, none did by the end of the study. The majority of patients were completely anuric throughout the study. Although it was possible that the presence of even a minimal amount of urine output could have had an impact on study outcomes, this was not shown by the various analyses of residual renal function at baseline or as a time-related variable.
Peritoneal membrane transport status at baseline had no effect on outcome, unlike in previous studies of patients on CAPD (4,5,34). High transport status in CAPD results in poor UF with consequent fluid overload. With APD, there is little fluid reabsorption across the peritoneum as cycle length is short. Potential fluid reabsorption during the long-day dwell was avoided by the use of icodextrin (35). This suggests that these treatment effects could ameliorate the detrimental effect on UF inherent in high transport status.
No patients with low transport status were enrolled in the study, in part because of the guidelines for prescribing APD in the study, which specified that CAPD patients with low transport status who became anuric would not be transferred to APD. One can only speculate that APD patients with preexisting low transport status may have already transferred to hemodialysis because of inadequate dialysis as a result of progressive decline in residual renal function or that their peritoneal membranes became more permeable with time on dialysis as has been shown in a number of studies (34).
Although it seems that the dropout rate in this study was high, the 2-yr technique survival of 62% is similar to that of other multicenter studies (12,25). The main cause of transfer to hemodialysis was peritonitis (12% of all patients, 43% of transfers). Although this rate of technique failure seems to be more than expected, it is in agreement with other studies of long-term PD reported by Davies et al. (34), in which peritonitis accounted for 13 to 54% of technique failure. A peritonitis rate of one episode in 18.8 patient-months may seem to be high, but this rate has been reported by others (36) and does fall within the UK Renal Association guidelines (2). It has been suggested that the outcome of peritonitis is adversely affected by length of time on PD (37), but the cure rate in this study was 81%, which compares favorably to the UK Renal Association standard of 80%.
A striking finding from this study is the association between poor UF and reduced survival. Baseline UF rate, unlike Ccrea, was shown to be a significant predictor of survival by both univariate and multivariate analyses. An important feature of the study design was a preestablished plan to manage patients’ dialysis prescriptions to prospectively determined targets for UF and Ccrea. There is clear evidence from the longitudinal analysis of these parameters, particularly during the first 6 mo of treatment optimization, that clinicians took appropriate steps to reach targets. Greater success was achieved in reaching the target for Ccrea than for UF, and at least part of the apparent improvement in the latter was due to informative censoring, as the dropout rate as a result of death in this group was significant.
To account for treatment changes, Cox modeling was also undertaken in which UF and Ccrea were treated as time-dependent covariates. Although not statistically significant, higher time-averaged UF continued to be associated with better survival. This, combined with the consistently poor UF achieved by patients who were below target at baseline and their poor UF capacity on the baseline PET, suggests a persistent problem of fluid removal in these patients. Many aspects of membrane function contribute to the achieved UF. The best documented is solute transport, which reduces UF by increasing the rate of glucose absorption and thus decreasing the osmotic gradient. However, solute transport accounts only for approximately 15% of the variance in achieved UF, other factors being fluid reabsorption and those determining osmotic conductance of the membrane—literally the amount of UF obtained for a given osmotic gradient. Factors that affect the osmotic conductance will include the hydraulic conductance of the membrane, membrane area, and the efficiency of various pore systems that determine the reflection coefficient for glucose. By using APD and icodextrin, the intention is to eradicate the bad effect of high transport on UF, as seems to have happened in this study. However, all of the other factors remain and indeed now assume relatively more importance as this study shows. These factors are difficult to measure individually but are lumped together as the UF capacity of the membrane—measured in the PET. Thus, the lower UF capacity in the patients below target is very significant and explains why these patients failed to achieve targets despite equivalent glucose use by 6 mo.
The reason that these patients did so poorly cannot be determined with certainty from this study, as fluid status in an individual patient is determined as much by fluid intake as by UF achieved (38). However, prolonged abnormal fluid status is one possible mechanism for poor survival, perhaps contributing to cardiovascular death. Equally, if these patients were euvolemic, then very low intakes of both fluid and salt would be required. Typically, <750 ml of UF in APD is associated with <50 mmol of sodium removal, as a result of sieving in short exchanges. Whatever the explanation, this study suggests that clinicians need to pay close attention to APD patients who consistently achieve low UF volumes. This observation is in keeping with those from other recent studies, such as Ates et al. (39), which suggested that the impact of loss of residual renal function on survival was due to its effect on salt and water removal and not the loss of small solute clearance.
In conclusion, the survival of anuric patients on APD is similar to other patients on PD. Furthermore, the poor prognosis of CAPD patients with high peritoneal solute transport does not apply to patients on APD. By using APD, it is possible to achieve sufficient small solute clearance and UF to treat successfully even anuric patients. This is important not just for patients who are already on PD but also for patients who are on hemodialysis, particularly those who no longer have vascular access. Finally, this study demonstrates the important role of UF in contrast to small solute clearance as an outcome parameter in patients on PD.
|
Appendix: Members of EAPOS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
|
Acknowledgments |
|---|
The results of the study have been presented in part at EuroPD, Brussels, May 2002, and the American Society of Nephrology, Philadelhia, October 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionAppendix: Members of EAPOSReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. van den Berg, T. T. Cnossen, C. J. A. M. Konings, J. P. Kooman, F. M. van der Sande, and K. M. L. Leunissen Different treatment options in peritoneal dialysis NDT Plus, October 1, 2008; 1(suppl_4): iv14 - iv17. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-H. Oh, J.-Y. Moon, J. Oh, S. G. Kim, Y.-H. Hwang, S. Kim, J. S. Lee, and C. Ahn Baseline peritoneal solute transport rate is not associated with markers of systemic inflammation or comorbidity in incident Korean peritoneal dialysis patients Nephrol. Dial. Transplant., July 1, 2008; 23(7): 2356 - 2364. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Baboolal, P. McEwan, S. Sondhi, P. Spiewanowski, J. Wechowski, and K. Wilson The cost of renal dialysis in a UK setting--a multicentre study Nephrol. Dial. Transplant., June 1, 2008; 23(6): 1982 - 1989. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. v. Biesen, N. Veys, N. Lameire, and R. Vanholder Why less success of the peritoneal dialysis programmes in Europe? Nephrol. Dial. Transplant., May 1, 2008; 23(5): 1478 - 1481. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Paniagua, D. Amato, S. Mujais, E. Vonesh, A. Ramos, R. Correa-Rotter, and W. H. Horl Predictive Value of Brain Natriuretic Peptides in Patients on Peritoneal Dialysis: Results from the ADEMEX Trial Clin. J. Am. Soc. Nephrol., March 1, 2008; 3(2): 407 - 415. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Khawar, K. Kalantar-Zadeh, W. K. Lo, D. Johnson, and R. Mehrotra Is the Declining Use of Long-Term Peritoneal Dialysis Justified by Outcome Data? Clin. J. Am. Soc. Nephrol., November 1, 2007; 2(6): 1317 - 1328. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. J. Wiggins, S. P. McDonald, F. G. Brown, J. B. Rosman, and D. W. Johnson High membrane transport status on peritoneal dialysis is not associated with reduced survival following transfer to haemodialysis Nephrol. Dial. Transplant., October 1, 2007; 22(10): 3005 - 3012. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. S. Rabindranath, J. Adams, T. Z. Ali, C. Daly, L. Vale, and A. M. MacLeod Automated vs continuous ambulatory peritoneal dialysis: a systematic review of randomized controlled trials Nephrol. Dial. Transplant., October 1, 2007; 22(10): 2991 - 2998. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Brimble, M. Walker, P. J. Margetts, K. K. Kundhal, and C. G. Rabbat Meta-Analysis: Peritoneal Membrane Transport, Mortality, and Technique Failure in Peritoneal Dialysis J. Am. Soc. Nephrol., September 1, 2006; 17(9): 2591 - 2598. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Rodrigues, M. Almeida, I. Fonseca, M. Martins, M. J. Carvalho, F. Silva, C. Correia, M. J. Santos, and A. Cabrita Peritoneal fast transport in incident peritoneal dialysis patients is not consistently associated with systemic inflammation Nephrol. Dial. Transplant., March 1, 2006; 21(3): 763 - 769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Van Biesen, A. Van Der Tol, N. Veys, C. Dequidt, D. Vijt, N. Lameire, and R. Vanholder The Personal Dialysis Capacity Test Is Superior to the Peritoneal Equilibration Test to Discriminate Inflammation as the Cause of Fast Transport Status in Peritoneal Dialysis Patients Clin. J. Am. Soc. Nephrol., March 1, 2006; 1(2): 269 - 274. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Rumpsfeld, S. P. McDonald, and D. W. Johnson Higher Peritoneal Transport Status Is Associated with Higher Mortality and Technique Failure in the Australian and New Zealand Peritoneal Dialysis Patient Populations J. Am. Soc. Nephrol., January 1, 2006; 17(1): 271 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. A. Brown Peritonitis: limiting the damage Nephrol. Dial. Transplant., August 1, 2005; 20(8): 1539 - 1541. [Full Text] [PDF]
|
|
|
|
|
|
S. H. Chung, O. Heimburger, B. Lindholm, and H. B. Lee Peritoneal dialysis patient survival: a comparison between a Swedish and a Korean centre Nephrol. Dial. Transplant., June 1, 2005; 20(6): 1207 - 1213. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Aanen, D. Venturoli, and S. J. Davies A detailed analysis of sodium removal by peritoneal dialysis: comparison with predictions from the three-pore model of membrane function Nephrol. Dial. Transplant., June 1, 2005; 20(6): 1192 - 1200. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. F. Flessner The transport barrier in intraperitoneal therapy Am J Physiol Renal Physiol, March 1, 2005; 288(3): F433 - F442. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Davies Are PD patients with or without residual renal function qualitatively different--or are they simply at different stages of the continuum of progressive uraemia? Nephrol. Dial. Transplant., February 1, 2005; 20(2): 270 - 272. [ |
) and <750 ml (group 2; 
) UF at baseline. (*between-group difference, P < 0.007; 
), pure technique (---) and combined patient and technique survival (---).