Abstract
Key Points
Nurse assistance is associated with a lower risk of transfer to hemodialysis for dialysis inadequacy after 6 months and for infection in the first 18 months.
Compared with automated peritoneal dialysis (PD), continuous ambulatory PD is associated with a higher risk of transfer to hemodialysis for mechanical issue during the first 18 months.
Suboptimal starters have a higher risk of transfer to hemodialysis due to psychosocial challenges in the first 6 months of PD.
Background
The end of peritoneal dialysis (PD) can be marked by kidney transplantation, death, or transfer to hemodialysis. We compared the risks of the different reasons for transfer to hemodialysis in patients on PD according to the use of assistance for PD care, PD modality, and the suboptimal starter status.
Methods
This was a retrospective study using data from the French Language PD Registry from patients who started PD between January 1, 2002, and December 31, 2018. We used Cox and Fine–Gray survival models to evaluate the risks of transfer to hemodialysis due to PD inadequacy, infection, mechanical issue, psychosocial issue, other PD-related causes, and other non–PD-related causes. Models were evaluated for three periods of PD vintage: 0–6 months, 6–18 months, and after 18 months.
Results
The study included 15,974 patients on incident PD treated in 170 French PD units. There were 6835 deaths, 5108 transfers to hemodialysis, and 3092 renal transplantations. Nurse-assisted PD was associated with a lower risk of transfer to hemodialysis for infection in the first 18 months (cause-specific hazard ratio [cs-HR], 0.51; 95% confidence interval [CI], 0.31 to 0.83 before 6 months) and for adequacy issues after 6 months (cs-HR, 0.59; 95% CI, 0.51 to 0.70 after 18 months). The risk of transfer for mechanical issue was higher in continuous ambulatory PD compared with automated PD during the first 18 months (cs-HR, 1.41; 95% CI, 1.00 to 1.99 before 6 months), but continuous ambulatory PD was associated with a lower risk of adequacy, infectious, or mechanical issue after 18 months. Finally, suboptimal starters have a higher risk of transfer due to psychosocial challenges in the first 6 months (cs-HR, 1.70; 95% CI, 1.03 to 2.81).
Conclusions
Distinct factors are associated with the risk of transfer from PD to in-center hemodialysis, according to the cause of the transfer. Some preventive measures targeting these risk factors may help to maintain patients in PD.
Podcast
This article contains a podcast at https://dts.podtrac.com/redirect.mp3/www.asn-online.org/media/podcast/K360/2025_03_27_KID0000000732.mp3
Keywords: clinical epidemiology, dialysis, dialysis access, dialysis volume, ESKD, epidemiology and outcomes, peritoneal dialysis
Visual Abstract
Introduction
Globally, the number of patients with ESKD requiring KRT is increasing, becoming a worldwide challenge.1–3 Nearly three million of these patients are treated with dialysis, a number expected to reach five million by 2027.4 The two main modalities of KRT are hemodialysis and peritoneal dialysis (PD).5–7 With the ongoing rise in the global burden of CKD, PD is often advocated as the preferred initial dialysis modality.8,9 It has notable advantages, such as better short-term survival compared with hemodialysis,10 improved quality of life, reduced financial burden, and the ability for patients to maintain their autonomy by undergoing treatment at home.5,7,11–14 PD is also recognized for its efficacy in many countries,15–18 providing a viable alternative when kidney transplantation, the most effective form of KRT is not possible.6
However, despite these advantages, PD remains underutilized globally, involving only approximately 11% of patients on chronic dialysis11,13 and 9% of all KRTs worldwide.13 Switching from one dialysis modality to another can occur for various reasons, including technical aspects specific to the currently used modality.6
The most frequently reported causes of transfer from PD to hemodialysis are infections, particularly PD-associated peritonitis, inadequate dialysis dose or difficulties in achieving adequate ultrafiltration, mechanical problems such as catheter dysfunction and dialysate leaks, and psychosocial issues.17,18 The incidence of the different causes of transfers to hemodialysis is not constant over time spent in PD. In the first 3–6 months after PD start, mechanical problems are the most frequent cause of transfer, whereas after this period, inadequacy and infections are the main causes of transfer.9,17,19 Therefore, one can hypothesize that the factors associated with the risk of transfer may be different according to the cause of the transfer. To anticipate the risk of transfer to hemodialysis and to implement individualized preventive strategies, it seems of great interest to better understand the risk factors associated with each different situation leading to transfer to hemodialysis. The aim of this study was to evaluate the factors associated with the different causes of transfer to in-center hemodialysis.
Methods
Study Population
This retrospective study involved patients older than 18 years who initiated PD between January 1, 2002, and December 31, 2018. Data were extracted from the French Language Peritoneal Dialysis Registry (RDPLF).20 The end of the study period was December 31, 2021.
Definition of Variables
The patient's characteristics were age at onset of PD; sex; diabetic status; modified Charlson Comorbidity Index (CCI); suboptimal starter status, defined as starting of PD after a period of <30 days on hemodialysis before PD21; treatment before PD, either hemodialysis or transplantation; PD modality—automated PD (APD) or continuous ambulatory PD (CAPD); inscription on the waiting list for kidney transplant; nephropathy; and assistance by nurse or family member for PD care. Modification of the CCI involved subtracting age underscoring from the CCI. Center-specific characteristics were the administrative type of the center and center experience estimated by the number of incident patients per year dichotomized with a cutoff of ten new patients per year by center.
Events of Interest and Competing Events
The event of interest was the transfer to facility-based hemodialysis because of a given cause. The reason for transfer is a declarative variable collected in the registry. The six categories are infection, inadequate dialysis (inadequate solute clearance associated with uremic syndrome or inadequate ultrafiltration volume associated with overhydration or poor nutrition), mechanical issue due to PD catheter malfunction, psychosocial issues (patient preference, loss of family assistance, and geographic isolation), other PD-related causes, and other non–PD-related causes. For a given cause of transfer to facility-based hemodialysis, the competing events considered were the five other potential causes of transfer and death or kidney transplantation.
Statistical Analyses
Categorical variables were described by their absolute numbers and percentages, while continuous variables were described by their median and interquartile ranges. Survival analyses were performed to investigate the associations between the risks of transfer to facility-based hemodialysis for any of the six registered causes and exposures variables. We constructed directed acyclic graphs (DAGs) to identify confounders for the associations between the three exposure variables—assistance, PD modality, and suboptimal starter—and the outcome transfer to hemodialysis for one of the six causes (Supplemental Figure 1). Confounders were identified through clinical expertise, prior evidence, and DAG analysis using the DAGitty R package.22 This method allowed to focus on a minimal sufficient adjustment set to address confounding while avoiding over adjustment. Age, sex, diabetes, CCI, and center size were selected as adjusting covariates.
Proportional risks survival models were considered to best account for the existence of competing risks. Cox survival models, which provide cause-specific hazard ratio (cs-HR), and Fine–Gray models, which estimate subdistribution hazard ratio (sd-HR), were used to assess the specific association between covariates and the event of interest. The cs-HR reflects the instantaneous risk of an event conditional on remaining event free, while the sd-HR accounts for the instantaneous risk in individuals who are either event-free or have experienced a competing event. This helps to interpret the effect of informative censoring due to differences in rate of mortality or transplantation. Given our focus on exploring etiologies and risk factors, we considered cs-HR as the primary result. However, reporting sd-HR alongside cs-HR offers additional insight into the effect of competing events, aligning with expert recommendations for survival analysis.23–25 The uncertainty of the results was expressed by the 95% confidence interval (CI). The proportional hazards assumption was assessed graphically and using Schoenfeld residuals. Owing to violation of this assumption, indicated by crossings in the cumulative incidence functions, we evaluated HRs for three PD vintage periods: <6 months (early), 6–18 months (middle), and more than 18 months (late). The decision was made not to include interaction terms to maintain clarity and interpretability of the primary results. The analysis was performed on all included patients on incident PD. Missing data were present in 2.7% of patients and were assumed to be missing completely at random. Therefore, we performed a complete case analysis. Statistical analyses were conducted using R 4.3.3 (R Foundation for Statistical Computing, Vienna, Austria).
The RDPLF has the approval of the French National Ethics Committee (Commission nationale de l'informatique et des libertés). This study took place within the framework of this authorization. This study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.26
Results
Patient Characteristics
A total of 16,042 patients on PD treated in 170 PD units were included in the study. Sixty-eight patients were excluded because of missing data concerning the specific reason for transfer or inconsistent data; they were excluded from the initial dataset. This left 15,974 patients for analysis, with a total follow-up time of 32,434 patient-years. Among these patients, 5108 (32%) were transferred to hemodialysis, 3092 (19%) received transplants, 6835 (43%) died, and 199 (1%) had kidney function recovery (Figure 1A). Seven hundred forty patients were still treated with PD at the end of the study (Figure 2).
Figure 1.
Cumulative incidence for PD cessation and its causes over time. (A) Cumulative incidence functions for the four causes of PD cessation. (B) Cumulative incidence functions for the six causes of transfer to hemodialysis. The three PD vintage periods are shown in different colors (<6 months, 6–18 months, and more than 18 months). HD, hemodialysis; PD, peritoneal dialysis.
Figure 2.
Flow chart. RDPLF, French Language Peritoneal Dialysis Registry.
The sociodemographic and clinical characteristics of this population are presented in Table 1. Among these patients, the mean age was 66.7 years (±17.2), 60% were male, and 80% had not received any treatment before admission to PD. By contrast, 9% of patients were suboptimal starters. 32% of these patients had significant comorbidity with a CCI ≥5. Vascular nephropathy was the most prevalent cause of kidney failure, present in 5102 (32%) patients. Seven thousand seven hundred fifty-nine (49%) patients were treated in general hospitals. Nurse assistance for PD care was provided to 7014 (44%) patients, and 9886 (61.9%) patients were treated with APD. There were 11,608 (73%) patients who were treated in small centers, and 4162 (26%) were on a kidney transplant waiting list.
Table 1.
Characteristics of the population
| Characteristics | All Patients (n=15,974) | Still on PD (n=740) | Death (n=6835) | Transplantation (n=3092) | Kidney Function Recovery (n=199) | Causes of Transfer | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Inadequate Dialysis (n=2024) | Other PD-Related Causes (n=591) | Other Non–PD-Related Causes (n=760) | Infection (n=825) | Mechanical (n=477) | Psychosocial Issues (n=431) | ||||||
| Age at PD initiation, mean (SD) | |||||||||||
| Median (IQR), yr | 66.7 (17.2) | 65.7 (16.1) | 77.1 (10.5) | 49.3 (14.2) | 67.9 (16.7) | 61.5 (16.6) | 65.2 (16.4) | 65.1 (15.4) | 63.6 (16.2) | 65.9 (16.5) | 66.0 (16.3) |
| Sex, No. (%) | |||||||||||
| Male | 9597 (60.1) | 385 (52.0) | 4085 (59.8) | 1853 (59.9) | 124 (62.3) | 1278 (63.1) | 303 (51.3) | 454 (59.7) | 507 (61.5) | 322 (67.5) | 286 (66.4) |
| Treatment before PD, No. (%) | |||||||||||
| No KRT | 12,770 (79.9) | 588 (79.5) | 5629 (82.4) | 2504 (81.0) | 138 (69.3) | 1513 (74.8) | 478 (80.9) | 576 (75.8) | 647 (78.4) | 370 (77.6) | 327 (75.9) |
| Hemodialysis | 2722 (17.0) | 129 (17.4) | 1138 (16.6) | 434 (14.0) | 59 (29.6) | 394 (19.5) | 89 (15.1) | 154 (20.3) | 138 (16.7) | 95 (19.9) | 92 (21.3) |
| Transplantation | 482 (3.0) | 23 (3.1) | 68 (1.0) | 154 (5.0) | 2 (1.0) | 117 (5.8) | 24 (4.1) | 30 (3.9) | 40 (4.8) | 12 (2.5) | 12 (2.8) |
| Suboptimal starter, No. (%) | |||||||||||
| Yes | 1471 (9.2) | 57 (7.7) | 692 (10.1) | 282 (9.1) | 19 (9.5) | 158 (7.8) | 42 (7.1) | 75 (9.9) | 72 (8.7) | 29 (6.1) | 45 (10.4) |
| Missing | 25 (0.2) | 0 (0) | 10 (0.1) | 7 (0.2) | 0 (0) | 2 (0.1) | 1 (0.2) | 2 (0.3) | 1 (0.1) | 0 (0) | 2 (0.5) |
| Modified CCI, No. (%) | |||||||||||
| 2 | 4938 (30.9) | 239 (32.3) | 953 (13.9) | 1990 (64.4) | 56 (28.1) | 689 (34.0) | 190 (32.1) | 255 (33.6) | 281 (34.1) | 159 (33.3) | 126 (29.2) |
| 3 | 2955 (18.5) | 168 (22.7) | 1255 (18.4) | 513 (16.6) | 44 (22.1) | 402 (19.9) | 126 (21.3) | 133 (17.5) | 166 (20.1) | 64 (13.4) | 84 (19.5) |
| 4 | 2609 (16.3) | 130 (17.6) | 1343 (19.6) | 266 (8.6) | 37 (18.6) | 293 (14.5) | 104 (17.6) | 124 (16.3) | 145 (17.6) | 93 (19.5) | 74 (17.2) |
| ≥5 | 5079 (31.8) | 192 (25.9) | 3102 (45.4) | 259 (8.4) | 57 (28.6) | 602 (29.7) | 157 (26.6) | 233 (30.7) | 208 (25.2) | 143 (30.0) | 126 (29.2) |
| Missing | 393 (2.5) | 11 (1.5) | 182 (2.7) | 64 (2.1) | 5 (2.5) | 38 (1.9) | 14 (2.4) | 15 (2.0) | 25 (3.0) | 18 (3.8) | 21 (4.9) |
| Nephropathy, No. (%) | |||||||||||
| PKD | 1017 (6.4) | 52 (7.0) | 147 (2.2) | 453 (14.7) | 4 (2.0) | 134 (6.6) | 56 (9.5) | 60 (7.9) | 59 (7.2) | 29 (6.1) | 23 (5.3) |
| Other | 474 (3.0) | 23 (3.1) | 177 (2.6) | 111 (3.6) | 15 (7.5) | 52 (2.6) | 18 (3.0) | 26 (3.4) | 21 (2.5) | 16 (3.4) | 15 (3.5) |
| Diabetic | 3089 (19.3) | 150 (20.3) | 1621 (23.7) | 266 (8.6) | 19 (9.5) | 451 (22.3) | 104 (17.6) | 138 (18.2) | 153 (18.5) | 97 (20.3) | 90 (20.9) |
| Vascular | 5102 (31.9) | 207 (28.0) | 3009 (44.0) | 395 (12.8) | 92 (46.2) | 511 (25.2) | 167 (28.3) | 217 (28.6) | 215 (26.1) | 141 (29.6) | 148 (34.3) |
| GN | 2393 (15.0) | 99 (13.4) | 428 (6.3) | 931 (30.1) | 19 (9.5) | 406 (20.1) | 92 (15.6) | 119 (15.7) | 152 (18.4) | 81 (17.0) | 66 (15.3) |
| Unknown | 1856 (11.6) | 90 (12.2) | 852 (12.5) | 324 (10.5) | 18 (9.0) | 210 (10.4) | 63 (10.7) | 83 (10.9) | 115 (13.9) | 66 (13.8) | 35 (8.1) |
| TIN | 952 (6.0) | 46 (6.2) | 318 (4.7) | 269 (8.7) | 13 (6.5) | 113 (5.6) | 42 (7.1) | 57 (7.5) | 56 (6.8) | 16 (3.4) | 22 (5.1) |
| Systemic | 426 (2.7) | 26 (3.5) | 102 (1.5) | 123 (4.0) | 13 (6.5) | 66 (3.3) | 21 (3.6) | 30 (3.9) | 23 (2.8) | 10 (2.1) | 12 (2.8) |
| Urologic | 515 (3.2) | 42 (5.7) | 102 (1.5) | 197 (6.4) | 3 (1.5) | 71 (3.5) | 23 (3.9) | 24 (3.2) | 23 (2.8) | 16 (3.4) | 14 (3.2) |
| Missing | 150 (0.9) | 5 (0.7) | 79 (1.2) | 23 (0.7) | 3 (1.5) | 10 (0.5) | 5 (0.8) | 6 (0.8) | 8 (1.0) | 5 (1.0) | 6 (1.4) |
| Center type, No. (%) | |||||||||||
| University hospital | 3097 (19.4) | 143 (19.3) | 1234 (18.1) | 702 (22.7) | 31 (15.6) | 405 (20.0) | 108 (18.3) | 137 (18.0) | 168 (20.4) | 85 (17.8) | 84 (19.5) |
| Nonprofit | 3742 (23.4) | 159 (21.5) | 1489 (21.8) | 857 (27.7) | 51 (25.6) | 456 (22.5) | 151 (25.6) | 203 (26.7) | 167 (20.2) | 110 (23.1) | 99 (23.0) |
| General hospital | 7759 (48.6) | 373 (50.4) | 3491 (51.1) | 1324 (42.8) | 107 (53.8) | 979 (48.4) | 290 (49.1) | 332 (43.7) | 414 (50.2) | 251 (52.6) | 198 (45.9) |
| Private | 1366 (8.6) | 63 (8.5) | 617 (9.0) | 209 (6.8) | 10 (5.0) | 183 (9.0) | 42 (7.1) | 86 (11.3) | 75 (9.1) | 31 (6.5) | 50 (11.6) |
| Missing | 10 (0.1) | 2 (0.3) | 4 (0.1) | 0 (0) | 0 (0) | 1 (0.0) | 0 (0) | 2 (0.3) | 1 (0.1) | 0 (0) | 0 (0) |
| Nurse-assisted PD, No. (%) | |||||||||||
| Self | 7606 (47.6) | 406 (54.9) | 1219 (17.8) | 2853 (92.3) | 92 (46.2) | 1312 (64.8) | 327 (55.3) | 424 (55.8) | 507 (61.5) | 245 (51.4) | 221 (51.3) |
| Family | 1345 (8.4) | 59 (8.0) | 784 (11.5) | 86 (2.8) | 22 (11.1) | 167 (8.3) | 45 (7.6) | 55 (7.2) | 59 (7.2) | 33 (6.9) | 35 (8.1) |
| Nurse | 7014 (43.9) | 275 (37.2) | 4828 (70.6) | 150 (4.9) | 85 (42.7) | 545 (26.9) | 219 (37.1) | 281 (37.0) | 258 (31.3) | 199 (41.7) | 174 (40.4) |
| Missing | 9 (0.1) | 0 (0) | 4 (0.1) | 3 (0.1) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (0.1) | 0 (0) | 1 (0.2) |
| PD modality, No. (%) | |||||||||||
| APD | 6088 (38.1) | 268 (36.2) | 1242 (18.2) | 2160 (69.9) | 57 (28.6) | 1061 (52.4) | 243 (41.1) | 301 (39.6) | 396 (48.0) | 176 (36.9) | 184 (42.7) |
| CAPD | 9886 (61.9) | 472 (63.8) | 5593 (81.8) | 932 (30.1) | 142 (71.4) | 963 (47.6) | 348 (58.9) | 459 (60.4) | 429 (52.0) | 301 (63.1) | 247 (57.3) |
| Diabetes, No. (%) | 5265 (33.0) | 238 (32.2) | 2907 (42.5) | 391 (12.6) | 44 (22.1) | 686 (33.9) | 201 (34.0) | 249 (32.8) | 256 (31.0) | 151 (31.7) | 142 (32.9) |
| Missing | 10 (0.1) | 0 (0) | 4 (0.1) | 2 (0.1) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 3 (0.4) | 0 (0) | 1 (0.2) |
| Center size, No. (%) | |||||||||||
| Small | 11,608 (72.7) | 542 (73.2) | 4921 (72.0) | 2161 (69.9) | 138 (69.3) | 1539 (76.0) | 434 (73.4) | 582 (76.6) | 613 (74.3) | 361 (75.7) | 317 (73.5) |
| Large | 4366 (27.3) | 198 (26.8) | 1914 (28.0) | 931 (30.1) | 61 (30.7) | 485 (24.0) | 157 (26.6) | 178 (23.4) | 212 (25.7) | 116 (24.3) | 114 (26.5) |
| Waiting list, No. (%) | 4162 (26.1) | 202 (27.3) | 170 (2.5) | 2705 (87.5) | 18 (9.0) | 549 (27.1) | 100 (16.9) | 128 (16.8) | 149 (18.1) | 80 (16.8) | 61 (14.2) |
| Missing | 533 (3.3) | 15 (2.0) | 255 (3.7) | 78 (2.5) | 18 (9.0) | 41 (2.0) | 14 (2.4) | 31 (4.1) | 35 (4.2) | 20 (4.2) | 26 (6.0) |
Waiting list: awaiting renal transplantation. A center is defined as small if the number of patients on incidental peritoneal dialysis is <10/yr. APD, automated peritoneal dialysis; CAPD, continuous ambulatory peritoneal dialysis; CCI, Charlson Comorbidity Index; IQR, interquartile range; PD, peritoneal dialysis; PKD, polycystic kidney disease; TIN, tubulointerstitial nephropathy.
During the study period, 5108 patients experienced transfer to hemodialysis. The cause of transfer was inadequacy in 2024 (40%) of these patients, infection in 825 (16%), mechanical in 477 (9%), and psychosocial in 431 (8%). There were 591 (12%) patients who transferred for other PD-related causes and 760 (15%) for other non–PD-related causes (Figure 1B). Patient characteristics did not differ between the six groups of causes of transfer to hemodialysis (Table 1).
Multivariate Analysis
Transfer to Hemodialysis for Adequacy Issues
Nurse assistance was associated with a lower risk of transfer to hemodialysis due to adequacy issues after 6 months of PD (cs-HR, 0.66; 95% CI, 0.50 to 0.87 and sd-HR, 0.57; 95% CI, 0.44 to 0.74 from 6 to 18 months and cs-HR, 0.59; 95% CI, 0.51 to 0.70 and sd-HR, 0.54; 95% CI, 0.45 to 0.63 beyond 18 months; Figure 3). CAPD was associated with a lower risk of transfer for adequacy issue after 18 months of PD in the Cox model (cs-HR, 0.78; 95% CI, 0.69 to 0.89) and during all periods in the Fine–Gray model (sd-HR, 0.64; 95% CI, 0.50 to 0.84 before 6 months, and sd-HR, 0.71; 95% CI, 0.59 to 0.87 from 6 to 18 months, and sd-HR, 0.77; 95% CI, 0.68 to 0.88 after 18 months). Suboptimal starters had a lower risk of transfer for adequacy issues after 18 months in the Fine–Gray model (sd-HR, 0.78; 95% CI, 0.62 to 0.99), with no significant differences observed in the Cox models across all periods.
Figure 3.
Estimates from the Cox and Fine–Gray models for risk of transfer to hemodialysis due to inadequate dialysis. Survival models adjusted on age, sex, diabetes, CCI, and center size. APD, automated peritoneal dialysis; CCI, Charlson Comorbidity Index; CI, confidence interval; cs-HR: cause-specific hazard ratio; sd-HR, subdistribution hazard ratio.
Transfer to Hemodialysis for Infection Issues
Figure 4 illustrates the results for transfer to hemodialysis due to infections (Figure 4). Family assistance was associated with a reduced risk of transfer for infection during the first 6 months (cs-HR, 0.20; 95% CI, 0.06 to 0.66 and sd-HR, 0.12; 95% CI, 0.04 to 0.39), although no significant association was found after 6 months. Nurse assistance was also associated with a reduced risk during the first 18 months (cs-HR, 0.51; 95% CI, 0.31 to 0.83 and sd-HR, 0.38; 95% CI, 0.24 to 0.61 in the first 6 months, and then cs-HR, 0.57; 95% CI, 0.39 to 0.85 and sd-HR, 0.53; 95% CI, 0.35 to 0.80 for 6–18 months). The risk remained lower after 18 months only according to the Fine–Gray model (sd-HR, 0.65; 95% CI, 0.50 to 0.85). No significant difference was observed for suboptimal starters compared with regular PD starters.
Figure 4.
Estimates from the Cox and Fine–Gray models for risk of transfer to hemodialysis due to psychosocial causes. Survival models adjusted on age, sex, diabetes, CCI, and center size.
Transfer to Hemodialysis for Mechanical Issues
Figure 5 highlights the multivariate Cox model results for transfers due to mechanical issues (Figure 5). Family assistance was associated with a reduced risk of transfer for mechanical reasons only during the middle period of PD (cs-HR, 0.36; 95% CI, 0.14 to 0.91 and sd-HR, 0.31; 95% CI, 0.12 to 0.79). Nurse-assisted patients did not have a different risk of transfer for mechanical issues compared with autonomous patients. The association between PD modality and transfer risk for mechanical issues varied between models. CAPD was associated with a higher risk of transfer in the first 18 months of PD when considering the Cox models, but the difference was NS considering the Fine–Gray models (cs-HR, 1.41; 95% CI, 1.00 to 1.99 and sd-HR, 1.12; 95% CI, 0.80 to 1.58 before 6 months and cs-HR, 1.49; 95% CI, 1.00 to 2.22 and sd-HR, 1.26; 95% CI, 0.84 to 1.89 from 6 to 18 months). Suboptimal starters had a lower risk during the middle period in the Fine–Gray model (sd-HR, 0.37; 95% CI, 0.15 to 0.92).
Figure 5.
Estimates from the Cox and Fine–Gray models for risk of transfer to hemodialysis due to infections. Survival models adjusted on age, sex, diabetes, CCI, and center size.
Transfer to Hemodialysis for Psychosocial Issues
The multivariate Cox model results for psychosocial challenges are shown in Figure 6. Family-assisted patients had a higher risk of transfer for psychosocial reasons in the first 6 months according to the Cox model (cs-HR, 2.20; 95% CI, 1.13 to 4.29), but no difference existed later on and considering the Fine–Gray models. Nurse-assisted patients had no different risk of transfer to hemodialysis for psychosocial causes compared with autonomous patients, except with the Fine–Gray model after 18 months of PD (sd-HR, 0.55; 95% CI, 0.34 to 0.89). PD modality was not associated with a significant difference in transfer risk for psychosocial reasons across the three time periods. Suboptimal starters demonstrated a higher risk of transfer during the early period, as shown by the Cox model (cs-HR, 1.70; 95% CI, 1.03 to 2.81), but the association was NS in later periods.
Figure 6.
Estimates from the Cox and Fine–Gray models for risk of transfer to hemodialysis due to mechanical issues. Survival models adjusted on age, sex, diabetes, CCI, and center size.
Transfer to Hemodialysis for Other PD-Related Causes
The estimates from the survival models evaluating the risk of transfer to hemodialysis due to other PD-related causes are presented in Supplemental Figure 2. Patients receiving family assistance exhibited a lower risk during the late period (cs-HR, 0.38; 95% CI, 0.18 to 0.83 and sd-HR, 0.32; 95% CI, 0.15 to 0.73), as well as those with nurse assistance (cs-HR, 0.58; 95% CI, 0.38 to 0.87 and sd-HR, 0.53; 95% CI, 0.34 to 0.80). No difference was seen according to the modality of PD for the risk of transfer to hemodialysis due to other PD-related causes. Neither difference was found concerning the suboptimal starters compared with regular starters.
Transfer to Hemodialysis for Other Non–PD-Related Causes
The results of the Cox and Fine–Gray models for the risk of transfer to hemodialysis due to other non–PD-related causes are shown in Supplemental Figure 3. Neither family nor nurse assistance was associated with a significantly different risk of transfer to hemodialysis for other causes not related to PD compared with patients on autonomous PD, according to the Cox models. Patients treated with CAPD had a higher risk of transfer in the first 6 months, only when considering the Cox model (cs-HR, 1.48; 95% CI, 1.05 to 2.08 and sd-HR, 1.08; 95% CI, 0.76 to 1.53). Suboptimal starters did not have a significantly different risk compared with regular starters, whatever the period.
Discussion
In this observational registry-based study of French patients undergoing PD, we identified that risk factors for transfer from PD to hemodialysis varied on the basis of the underlying cause. Nurse-assisted PD was significantly associated with a reduced risk of transfer due to inadequacy after 6 months of PD, but not in the earlier period. Conversely, the risk was lower in the first 18 months for the risk of transfer due to infection issues. CAPD consistently showed a lower risk of transfer for adequacy, infection, and mechanical causes after 18 months, but the risk of transfer for mechanical reasons was higher in the first 18 months in the Cox model only. Suboptimal starters demonstrated a higher risk of transfer due to psychosocial causes in the first 6 months considering the Cox model. A summary of the results of the different models proposed in our work is presented in Supplemental Table 1.
The end of PD can be marked by kidney transplantation, death, or transfer to hemodialysis.17 One of the main challenges associated with PD is its limited technique survival, which often necessitates a transfer to center-based hemodialysis.17,27 The median time to transfer to hemodialysis ranged from 2 to 2.4 years, according to prior studies.9,28 In the international Peritoneal Dialysis Outcomes and Practice Patterns Study, 24%–35% of patients had hemodialysis transfer by 3 years on PD.29 The event of transferring to hemodialysis represents a significant adverse outcome in patients' lives.30 Understanding the mechanisms leading to this transfer is crucial for anticipating and mitigating this risk.
It has been described that the size of the center, presence of multiple comorbidities such as diabetes, and high body mass index are risk factors associated with the discontinuation of PD.8,29 However, these results were assessed independently of the specific causes for the transfer to hemodialysis.
We have shown that nurse-assisted PD was associated with a lower risk of transfer due to infections in the early period of PD and after only in the Fine–Gray model. Duquennoy et al. found that nurse-assisted care was associated with a reduced risk of peritonitis in the elderly.31 Furthermore, it has been shown that nurse-assisted PD had a protective effect against peritonitis in diabetic patients.32 A study conducted in Denmark also revealed that patients receiving assisted PD had a lower risk of infection.33 Conversely, a study conducted in Taiwan showed that elderly patients receiving assistance from a home caregiver or family member had the same risk of peritonitis as those on self-managed PD. This result can be explained by the fact that in Taiwan, assistance is primarily provided by foreign workers trained by PD centers who may not have the necessary prerequisites or experience to follow the rigorous PD protocols.34 Nurses play a crucial role in training patients and their families in realizing PD exchanges and to detect peritonitis earlier and prevent its episodes.35 Nurse-assisted PD contributes to maintaining a cleaner and safer PD technique, thereby reducing the risk of severe infections that could lead to a transfer to hemodialysis.
We also found that nurse-assisted PD had a protective effect after 6 months of PD against transfer due to dialysis inadequacy. The effect may take time to manifest because adequacy emerges as the main cause of transfer after several months of PD, as illustrated by the cumulative incidence function curves for the various causes of transfer (Figure 1B). This result is supported by previous work from researchers at the Nephrology Department in Caen focusing on the role of assisted PD, which revealed that nurse-assisted PD was associated with a lower risk of transfer to hemodialysis because of inadequate dialysis.17 Nurses can closely monitor patients on PD and promptly alert nephrologists if there is a need to adjust the PD prescription. The presence of nurses at home may enhance PD observance, potentially leading to improved PD adequacy, particularly over time as patients may experience treatment-related burnout.36 The delayed benefits of nurse assistance were recently shown in another study from our team concerning the global risk of transfer to hemodialysis.37
The increased risk of CAPD for transfer due to mechanical issues may be attributed to the advantage of tidal PD facilitated by APD. In patients with nonoptimal catheter drainage or those experiencing drain pain, reducing the draining time is believed to enable the continuation of the technique despite these mechanical dysfunctions.38 This finding may be partly explained by the demographics of patients on CAPD because older individuals are more likely to receive nurse-assisted CAPD, while younger, more independent patients are predominantly treated with APD. For example, in the 18- to 25-year age group, 68% are treated with APD and 32% with CAPD, whereas among patients older than 80 years, 12% receive APD compared with 88% on CAPD. Hemodialysis is often considered too burdensome for many elderly patients, prompting greater efforts to sustain them on PD until death. This focus could contribute to the reduced transfer rates to hemodialysis in this population. Consequently, the observed benefits of CAPD in the late period may reflect an age-related effect mediated by PD modality.
Interestingly, we found that suboptimal starters had an increased risk of transfer due to psychosocial challenges, during the first 6 months. A 2013 study using French data already concluded that suboptimal starters had a higher risk of transfer to hemodialysis.21 Interviews of patients with ESKD and their relatives revealed that the sense of security was an important factor influencing the choice of dialysis modality.39 Acute dialysis initiation can increase anxiety and reduce confidence in continuing with home therapy. This finding warrants further validation, as providing enhanced psychosocial support to suboptimal starters could prove to be a valuable approach.
There were cases of discordances between the estimates from the Cox and Fine–Gray models. The Cox proportional hazards model focuses solely on the cause-specific hazard, estimating the instantaneous risk of an event occurring for a specific cause while other competing events are censored. By contrast, in the Fine–Gray subdistribution hazard model, the probability of the event of interest is considered in the presence of competing events. These methodological differences have practical implications for interpreting results. The Fine–Gray model provides a perspective closer to the cumulative incidence function, which reflects the absolute risk of an event over time in a competing-risk framework. This can lead to differences in hazard ratios, especially when the proportion of competing events is significant (Supplemental Figure 4). It is therefore important to choose the appropriate analytical framework on the basis of the research question. The Cox model is useful for understanding the dynamics of a specific cause and is recommended for studying etiological questions, such as in this observational work. On the other hand, the Fine–Gray model offers insights into the overall probability of the event occurring in a competing-risk context and is advised for prognostic studies. Presentation of both models is recommended for an optimal understanding of the effect of risk factors on competing risks.24,25
Our study is based on a nationwide registry of patients treated with PD. We used robust statistical methods accounting for the presence of competitive risks and adjusted on patient-level characteristics. We were able to follow patients for up to 19 years, covering a substantial portion of their dialysis history. Finally, to the best of our knowledge, our study is the first to focus on the association of risk factors with the different causes of transfer to hemodialysis.
However, our results must be interpreted with certain limitations in mind. The observational design does not allow us to draw causal conclusions. Causes of transfer from PD to hemodialysis, such as adequacy issue, could be prone to a subjective assessment from the nephrologists, therefore different according to the center of treatment. The covariates in the survival models were not treated as time dependent, whereas their values may vary over time, influencing the association with the outcomes. The cumulative incidence functions for transfer due to inadequacy, infection, mechanical, and psychosocial reasons, and according to the three exposures, are shown in Supplemental Figure 5.
Several variables of interest were not included because they were not available in the RDPLF database. Information about the residual renal function, socioeconomic characteristics of the patients, and nutritional status are unavailable. These factors are known to influence the outcome of PD and could have affected the reported associations. Although causes of technical failure, such as peritonitis, inadequate PD, and mechanical and psychosocial reasons, were clearly coded in the database, other reasons for technical failure were less well described and lacked a standardized definition.
The main findings of our study are that nurse assistance is associated with a lower risk of transfer to hemodialysis for dialysis inadequacy after 6 months and with a lower risk of transfer for infection in the first 18 months. The risk of transfer for mechanical issue is higher in CAPD compared with APD during the first 18 months, but CAPD is associated with a lower risk for adequacy, infectious, or mechanical issues after 18 months. Finally, suboptimal starters have a higher risk of transfer due to psychosocial challenges. Recognizing these risk factors could help target specific patients with tailored preventive measures to reduce the risk of transfer for each particular cause. Further studies would be needed to explore the effect of such measures.
Supplementary Material
Acknowledgments
We would like to express our gratitude to all patients, nurses, and nephrologists in the renal units who provided data to the RDPLF.
Footnotes
See related editorial, “Strategies to Reduce Technique Failure in Peritoneal Dialysis,” on pages 496–497.
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/A903.
Funding
None.
Author Contributions
Conceptualization: Antoine Lanot, Thierry Lobbedez.
Data curation: Nanti E. Adoukonou.
Formal analysis: Nanti E. Adoukonou, Antoine Lanot.
Methodology: Antoine Lanot.
Supervision: Antoine Lanot.
Validation: Clémence Bechade, Annabel Boyer, Antoine Lanot, Thierry Lobbedez.
Writing – original draft: Nanti E. Adoukonou.
Writing – review & editing: Clémence Bechade, Annabel Boyer, Antoine Lanot, Thierry Lobbedez.
Data Sharing Statement
Partial restrictions to the data and/or materials apply. The data that support the findings of this study are available from the corresponding author upon reasonable request.
Supplemental Material
This article contains the following supplemental material online at http://links.lww.com/KN9/A902.
Supplemental Figure 1. DAG illustrating the a priori assumptions regarding the relationships between exposures, outcome of interest, and potential mediators and confounders. The generic outcome transfer to hemodialysis was used, and the depicted relationships were considered consistent across the six causes of transfer to hemodialysis.
Supplemental Figure 2. Factors associated with the risk of transfer to hemodialysis due to other PD-related causes in patients treated with PD.
Supplemental Figure 3. Factors associated with the risk of transfer to hemodialysis due to other non–PD-related causes in patients treated with PD.
Supplemental Figure 4. Handling of competing events according to Cox and Fine–Gray models.
Supplemental Figure 5. Cumulative incidence functions for transfer due to inadequacy, infection, mechanical, and psychosocial reasons and according to the three exposures.
Supplemental Table 1. Summary of the associations found in the different survival models.
References
- 1.Foreman KJ Marquez N Dolgert A, et al. Forecasting life expectancy, years of life lost, and all-cause and cause-specific mortality for 250 causes of death: reference and alternative scenarios for 2016-40 for 195 countries and territories. Lancet. 2018;392(10159):2052–2090. doi: 10.1016/S0140-6736(18)31694-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Liyanage T Ninomiya T Jha V, et al. Worldwide access to treatment for end-stage kidney disease: a systematic review. Lancet. 2015;385(9981):1975–1982. doi: 10.1016/S0140-6736(14)61601-9 [DOI] [PubMed] [Google Scholar]
- 3.Cho Y, See EJ, Htay H, Hawley CM, Johnson DW. Early peritoneal dialysis technique failure: review. Perit Dial Int. 2018;38(5):319–327. doi: 10.3747/pdi.2018.00017 [DOI] [PubMed] [Google Scholar]
- 4.Davenport A. Chronic Kidney Failure: Renal Replacement Therapy. Elsevier; 2019. https://www.sciencedirect.com/science/article/abs/pii/B9780323531863000036 [Google Scholar]
- 5.Jaar BG Plantinga LC Crews DC, et al. Timing, causes, predictors and prognosis of switching from peritoneal dialysis to hemodialysis: a prospective study. BMC Nephrol. 2009;10:3. doi: 10.1186/1471-2369-10-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gallieni M, Giordano A, Ricchiuto A, Gobatti D, Cariati M. Dialysis access: issues related to conversion from peritoneal dialysis to hemodialysis and vice versa. J Vasc Access. 2017;18(suppl 1):41–46. doi: 10.5301/jva.5000695 [DOI] [PubMed] [Google Scholar]
- 7.McDonald SP, Marshall MR, Johnson DW, Polkinghorne KR. Relationship between dialysis modality and mortality. J Am Soc Nephrol. 2009;20(1):155–163. doi: 10.1681/ASN.2007111188 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Piarulli P Vizzardi V Alberici F, et al. Peritoneal dialysis discontinuation: to the root of the problem. J Nephrol. 2023;36(7):1763–1776. doi: 10.1007/s40620-023-01759-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lan PG Clayton PA Johnson DW, et al. Duration of hemodialysis following peritoneal dialysis cessation in Australia and New Zealand: proposal for a standardized definition of technique failure. Perit Dial Int. 2016;36(6):623–630. doi: 10.3747/pdi.2015.00218 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lan PG, Clayton PA, Saunders J, Polkinghorne KR, Snelling PL. Predictors and outcomes of transfers from peritoneal dialysis to hemodialysis. Perit Dial Int. 2015;35(3):306–315. doi: 10.3747/pdi.2013.00030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jain AK, Blake P, Cordy P, Garg AX. Global trends in rates of peritoneal dialysis. J Am Soc Nephrol. 2012;23(3):533–544. doi: 10.1681/ASN.2011060607 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Okpechi IG Jha V Cho Y, et al. The case for increased peritoneal dialysis utilization in low- and lower-middle-income countries. Nephrology. 2022;27(5):391–403. doi: 10.1111/nep.14024 [DOI] [PubMed] [Google Scholar]
- 13.Bello AK Okpechi IG Osman MA, et al. Epidemiology of peritoneal dialysis outcomes. Nat Rev Nephrol. 2022;18(12):779–793. doi: 10.1038/s41581-022-00623-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Brown EA Johansson L Farrington K, et al. Broadening Options for Long-term Dialysis in the Elderly (BOLDE): differences in quality of life on peritoneal dialysis compared to haemodialysis for older patients. Nephrol Dial Transplant. 2010;25(11):3755–3763. doi: 10.1093/ndt/gfq212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lee H Manns B Taub K, et al. Cost analysis of ongoing care of patients with end-stage renal disease: the impact of dialysis modality and dialysis access. Am J Kidney Dis. 2002;40(3):611–622. doi: 10.1053/ajkd.2002.34924 [DOI] [PubMed] [Google Scholar]
- 16.Mehrotra R, Devuyst O, Davies SJ, Johnson DW. The current state of peritoneal dialysis. J Am Soc Nephrol. 2016;27(11):3238–3252. doi: 10.1681/ASN.2016010112 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lanot A, Bechade C, Boyer A, Ficheux M, Lobbedez T. Assisted peritoneal dialysis and transfer to haemodialysis: a cause-specific analysis with data from the RDPLF. Nephrol Dial Transplant. 2021;36(2):330–339. doi: 10.1093/ndt/gfaa289 [DOI] [PubMed] [Google Scholar]
- 18.Boyer A Lanot A Lambie M, et al. Trends in peritoneal dialysis technique survival, death, and transfer to hemodialysis: a decade of data from the RDPLF. Am J Nephrol. 2021;52(4):318–327. doi: 10.1159/000515472 [DOI] [PubMed] [Google Scholar]
- 19.Kolesnyk I, Dekker FW, Boeschoten EW, Krediet RT. Time-dependent reasons for peritoneal dialysis technique failure and mortality. Perit Dial Int. 2010;30(2):170–177. doi: 10.3747/pdi.2008.00277 [DOI] [PubMed] [Google Scholar]
- 20.Verger C Ryckelynck JP Duman M, et al. French peritoneal dialysis registry (RDPLF): outline and main results. Kidney Int. 2006;70(103):S12–S20. doi: 10.1038/sj.ki.5001911 [DOI] [PubMed] [Google Scholar]
- 21.Lobbedez T, Verger C, Ryckelynck JP, Fabre E, Evans D. Outcome of the sub-optimal dialysis starter on peritoneal dialysis. Report from the French Language Peritoneal Dialysis Registry (RDPLF). Nephrol Dial Transplant. 2013;28(5):1276–1283. doi: 10.1093/ndt/gft018 [DOI] [PubMed] [Google Scholar]
- 22.Textor J, van der Zander B, Gilthorpe MS, Liskiewicz M, Ellison GT. Robust causal inference using directed acyclic graphs: the R package 'dagitty'. Int J Epidemiol. 2016;45(6):1887–1894. doi: 10.1093/ije/dyw341 [DOI] [PubMed] [Google Scholar]
- 23.Austin PC, Fine JP. Practical recommendations for reporting Fine-Gray model analyses for competing risk data. Stat Med. 2017;36(27):4391–4400. doi: 10.1002/sim.7501 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lau B, Cole SR, Gange SJ. Competing risk regression models for epidemiologic data. Am J Epidemiol. 2009;170(2):244–256. doi: 10.1093/aje/kwp107 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Latouche A, Allignol A, Beyersmann J, Labopin M, Fine JP. A competing risks analysis should report results on all cause-specific hazards and cumulative incidence functions. J Clin Epidemiol. 2013;66(6):648–653. doi: 10.1016/j.jclinepi.2012.09.017 [DOI] [PubMed] [Google Scholar]
- 26.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP.; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344–349. doi: 10.1016/j.jclinepi.2007.11.008 [DOI] [PubMed] [Google Scholar]
- 27.Bonenkamp AA van Eck van der Sluijs A Dekker FW, et al. Technique failure in peritoneal dialysis: modifiable causes and patient-specific risk factors. Perit Dial Int. 2023;43(1):73–83. doi: 10.1177/08968608221077461 [DOI] [PubMed] [Google Scholar]
- 28.Perl J Wald R Bargman JM, et al. Changes in patient and technique survival over time among incident peritoneal dialysis patients in Canada. Clin J Am Soc Nephrol. 2012;7(7):1145–1154. doi: 10.2215/CJN.01480212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lambie M Zhao J McCullough K, et al. Variation in peritoneal dialysis time on therapy by country: results from the peritoneal dialysis outcomes and practice patterns study. Clin J Am Soc Nephrol. 2022;17(6):861–871. doi: 10.2215/CJN.16341221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Manera KE Johnson DW Craig JC, et al. Establishing a core outcome set for peritoneal dialysis: report of the SONG-PD (standardized outcomes in nephrology-peritoneal dialysis) consensus workshop. Am J Kidney Dis. 2020;75(3):404–412. doi: 10.1053/j.ajkd.2019.09.017 [DOI] [PubMed] [Google Scholar]
- 31.Duquennoy S, Béchade C, Verger C, Ficheux M, Ryckelynck JP, Lobbedez T. Is peritonitis risk increased in elderly patients on peritoneal dialysis? Report from the French Language Peritoneal dialysis registry (RDPLF). Perit Dial Int. 2016;36(3):291–296. doi: 10.3747/pdi.2014.00154 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Benabed A, Bechade C, Ficheux M, Verger C, Lobbedez T. Effect of assistance on peritonitis risk in diabetic patients treated by peritoneal dialysis: report from the French Language Peritoneal Dialysis Registry. Nephrol Dial Transplant. 2016;31(4):656–662. doi: 10.1093/ndt/gfw011 [DOI] [PubMed] [Google Scholar]
- 33.Povlsen JV, Ivarsen P. Assisted automated peritoneal dialysis (AAPD) for the functionally dependent and elderly patient. Perit Dial Int. 2005;25(suppl 3):S60–S63. doi: 10.1177/089686080502503s15 [DOI] [PubMed] [Google Scholar]
- 34.Cheng CH, Shu KH, Chuang YW, Huang ST, Chou MC, Chang HR. Clinical outcome of elderly peritoneal dialysis patients with assisted care in a single medical centre: a 25-year experience. Nephrology (Carlton). 2013;18(6):468–473. doi: 10.1111/nep.12090 [DOI] [PubMed] [Google Scholar]
- 35.Yang Z, Xu R, Zhuo M, Dong J. Advanced nursing experience is beneficial for lowering the peritonitis rate in patients on peritoneal dialysis. Perit Dial Int. 2012;32(1):60–66. doi: 10.3747/pdi.2010.00208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Oveyssi J Manera KE Baumgart A, et al. Patient and caregiver perspectives on burnout in peritoneal dialysis. Perit Dial Int. 2021;41(5):484–493. doi: 10.1177/0896860820970064 [DOI] [PubMed] [Google Scholar]
- 37.Boyer A, Lanot A, Ficheux M, Guillouet S, Bechade C, Lobbedez T. The time-dependent effect of assistance on peritoneal dialysis duration: an analysis of data from the French Language Peritoneal dialysis registry. Kidney360. 2024;5(10):1500–1509. doi: 10.34067/KID.0000000577 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Blake PG, Sloand JA, McMurray S, Jain AK, Matthews S. A multicenter survey of why and how tidal peritoneal dialysis (TPD) is being used. Perit Dial Int. 2014;34(4):458–460. doi: 10.3747/pdi.2013.00314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Lee A, Gudex C, Povlsen JV, Bonnevie B, Nielsen CP. Patients’ views regarding choice of dialysis modality. Nephrol Dial Transplant. 2008;23(12):3953–3959. doi: 10.1093/ndt/gfn365 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Partial restrictions to the data and/or materials apply. The data that support the findings of this study are available from the corresponding author upon reasonable request.