Observational Study Medicine ® Appropriateness of the dialysis modality selection process A cross-sectional study Karla Salas-Gama, MDa,b,c,d,* , Juan-Manuel Díaz-Gómez, PhDe,f, Ignasi Bolíbar Ribas, PhDd,g,h Abstract Studies that specifically quantify the appropriateness of the process of dialysis modality selection are lacking. Peritoneal dialysis (PD) offers clinical and social advantages over hemodialysis (HD), but may be underused. We aimed to determine the appropriateness of the process of dialysis modality selection and quantify the percentage of patients who could potentially have been PD candidates. We performed a cross-sectional study that included adult patients from a hospital Nephrology Department in Barcelona who started dialysis between 2014 and 2015. We assessed the appropriateness of dialysis modalities selection by defining 3 sequential domains based on 3 critical steps in choosing a dialysis modality: eligibility for either treatment, information about modalities, and shared decision-making. We obtained data using medical records and a patient questionnaire. The dialysis modality selection process was considered appropriate when patients had no contraindications for the selected option, received complete information about both modalities, and voluntarily chose the selected option. A total of 141 patients were included in this study. The median age was 72 years (interquartile range 63–82 years), and 65% of the patients were men. The dialysis modality selection process was potentially inappropriate in 22% of the participants because of problems related to information about dialysis modalities (15%) or shared decision-making (7%). Appropriate PD use can potentially increase from 17% to 38%. Patient age and lack of information regarding dialysis options were independently associated with the potential degree of inappropriate dialysis modality selection. Our findings indicate areas for improvement in the selection of dialysis modalities. With better education and shared decision-making, the number of patients with PD could potentially double. The analysis of appropriateness is a helpful approach for studying renal replacement treatment patterns and identifying strategies to optimize their use. Abbreviations: CKD = chronic kidney disease, ESKD = end-stage kidney disease, HD = hemodialysis, KT = kidney transplantation, PD = peritoneal dialysis, RRT = renal replacement therapies. Keywords: appropriateness, chronic kidney disease, peritoneal dialysis, renal replacement treatment, underuse. 1. Introduction renal replacement therapy (RRTs) along the patient pathway since switching between dialysis modalities is often essential.[11] Chronic kidney disease (CKD) is a global health problem with According to expert opinion,[12,13] PD should ideally reach 30% a worldwide prevalence of 9%.[1] Patients with end-stage kidney to 45% of patients starting RRT. However, the use of PD is disease (ESKD) represent 1% to 6% of this group, and their considerably low in clinical practice. Figures for Europe in 2019 treatment consumes 1% to 3% of all healthcare resources.[2,3] showed that 84% of patients with CKD started RRT with HD, Similar to hemodialysis (HD), peritoneal dialysis (PD) is an 5% underwent kidney transplantation (KT), and only 11% effective technique for the treatment of ESKD,[4–9] but most eco- started PD.[14] nomic evaluations indicate that it is more cost-effective.[2,3,10,11] Most patients with ESKD are potential candidates for Authors of a recent systematic review suggested a more com- either modality of dialysis.[15] The percentage of contrain- prehensive cost-effectiveness approach that considers different dications reported for PD varies between 17% and 28% of This research did not receive any specific grants from funding agencies in the (IIB Sant Pau), Barcelona, Spain, h Universitat Autònoma de Barcelona (UAB), public, commercial, or not-for-profit sectors. Barcelona, Spain. The authors have no conflicts of interest to disclose. *Correspondence: Karla Salas-Gama, Quality, Process and Innovation Direction, The datasets generated during and/or analyzed during the current study are Vall d’Hebron University Hospital, Passeig de la Vall d’Hebron, 119, 08035 available from the corresponding author on reasonable request. Barcelona, Spain (email: karla.salas@vallhebron.cat). Copyright © 2022 the Author(s). Published by Wolters Kluwer Health, Inc. Supplemental Digital Content is available for this article. This is an open-access article distributed under the terms of the Creative a Quality, Process and Innovation Direction, Valld’Hebron Hospital Universitari, Commons Attribution-Non Commercial License 4.0 (CCBY-NC), where it is Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain, b Health Services permissible to download, share, remix, transform, and buildup the work provided Research Group, Vall d’Hebron Institut de Recerca (VHIR), Vall d’Hebron it is properly cited. The work cannot be used commercially without permission Hospital Universitari, Vall d’Hebron Barcelona Hospital Campus, Barcelona, from the journal. Spain, c PhD candidate at the Methodology of Biomedical Research and Public How to cite this article: Salas-Gama K, Díaz-Gómez J-M, Bolíbar Ribas I. Health program, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain, Appropriateness of the dialysis modality selection process: A cross-sectional d Consortium for Biomedical Research in Epidemiology and Public Health- study. Medicine 2022;101:42(e31041). CIBERESP, Barcelona, Spain, e Nephrology Department, Fundación Puigvert, IIB Sant Pau, Barcelona, Spain, f Medicine Department, Universitat de Vic Received: 13 April 2022 / Received in final form: 1 September 2022 / Accepted: (UVic-UCC), Vic, Spain, g Department of Clinical Epidemiology and Public Health, 7 September 2022 Hospital de la Santa Creu i Sant Pau Biomedical Research Institute Sant Pau http://dx.doi.org/10.1097/MD.0000000000031041 1 Salas-Gama et al. • Medicine (2022) 101:42Medicine patients.[10,16] This variability may be explained by the fact about possible treatment options, and promptly prepare for HD that some authors include non-clinical contraindications for o peritoneal treatment. Not all patients can enter the program, PD in their calculations, whereas others consider such features due to their clinical (CKD with acute deterioration that did not as barriers, but no contraindications.[17–19] There is still insuf- have the opportunity to go through predialysis counseling or ficient high-level evidence to recommend one therapy over complete the whole counseling follow-up) or care (new ESKD another.[15,18,20] When different therapeutic options are effec- patients without previous visits at hospital) situation. tive, but their practical characteristics differ and could poten- The inclusion criteria were patients aged 18 years or older tially impact the patient’s life, choosing a modality should who started HD or PD between January 1, 2014 and December focus on patients’ values, preferences, and a voluntary and 31, 2015. Exclusion criteria were conditions or circumstances in informed choice.[15,21] which signed consent could not be obtained, such as cognitive Appropriate care refers to high-quality care by offering the deterioration, mental disorders, severe hearing loss, language right service to the right patient at the right time and using effec- barrier, death, and inability to contact the patient, explicitly liv- tive and efficient resources that are consistent with the values ing outside the hospital reference area. and preferences of individuals.[22]. However, the dimension of The average number of patients that initiated dialysis yearly the “right patient,” is not always considered in appropriateness at the nephrology department determined the sample size. We analysis, with a notable absence of variables related to the extent estimated a sample size of 143 participants based on an assumed of patient engagement in decision-making and decisional qual- percentage of appropriate use of 75%, confidence level of 95% ity.[23] When the decision-making process is facilitated, approxi- and margin of error of 5%. We assumed an 18% loss based on mately 50% of the candidates choose PD.[24,25] the annual mortality rate described for dialysis patients.[31] Although the underuse of PD has been discussed in previous We identified potential participants from the nephrology studies, none has explicitly quantified it.[26–30] This study aimed department register of incident patients on dialysis. We classi- to determine the appropriateness of dialysis modality selection fied the patients according to the dialysis modality performed process and quantify the percentage of patients who could on day 90, allowing them to get to the modality they wanted. All potentially have been PD candidates. participants voluntarily signed a written informed consent form before inclusion in the study and data collection. The study outcome was the appropriateness of dialysis modality selection. We assessed this, according to the values 2. Methods and preferences of individuals, by defining 3 sequential domains We performed a cross-sectional study to analyze the appropri- based on 3 critical steps in the pathway for choosing a dialy- ateness of dialysis modalities selection in the nephrology depart- sis modality: eligibility for either treatment, information about ment of a third-level university hospital with a reference area modalities, and shared decision-making (Table 1, Supplemental of 404,672 inhabitants in Barcelona, Spain. The nephrology Digital Content 2, http://links.lww.com/MD/H565).[19] The first department provides all RRT modalities, conservative man- domain was related to PD and HD eligibility assessment by look- agement, and a consultation unit in which nephrologists and ing for contraindications or barriers for both modalities already nurses specifically focus on the management of ESKD. When defined in the medical literature.[15–19] A research nephrologist patients with CKD reach a glomerular filtration rate <20 mL/ (KS), independent of the nephrology department, reviewed and min/1.73 m2 they are referred to the counseling and predialysis collected clinical and sociodemographic information from each education program performed entirely in the hospital setting. In participant’s clinical records. our health system there is no ESKD counseling in primary care The second domain was related to the degree of information or long-term care facilities outside of the hospital. The predial- provided for both modalities. We defined: complete informa- ysis education program does the necessary close follow-up to tion when the patient received information about either dialysis implement procedures and treatments for the control of the dis- modality and considered that this information was adequate, ease progression, such as control of blood pressure, lifestyle and incomplete information when only one modality was provided diet modifications, and censoring some drugs that could worsen to a patient potentially eligible for the 2 dialysis modalities, glomerular filtration rate. Patients will also receive information and absence of information when no information about the Table 1 Domains for assessing the appropriateness of the dialysis modality selection process. Domain 1: Assessment of the presence of the clinical contraindications or barriers to dialysis modalities (*) Contraindications or barriers for HD  1.Inability to secure vascular access  2.Presence of relative contraindications: poor cardiac condition, needle phobia, and coagulopathy Contraindications or barriers for PD  1. Presence of abdominal-peritoneal pathology.  2. Presence of relative contraindications (hiatal hernia with severe reflux esophagitis, severe diabetic gastroparesis, severe chronic obstructive pulmonary disease, severe malnutrition, giant polycystic kidney disease, severe spine disease, morbid obesity).  3. Inability to perform peritoneal dialysis at home: Problems in self-care: limitation in functional status (Karnofsky Scale <60) without the support of carers; Problems related to home characteristics: after an adequate process of information, the patient declares not to have the requirements at home to perform peritoneal dialysis. Domain 2: Degree of information received about both dialysis modalities  1. Complete information: the patient considered that the information received about either dialysis modalities (or information of one modality in those patients who only have the option for one of the 2 treatments) was: highly adequate, adequate, or neither adequate nor inadequate.  2. Incomplete information: only one modality was provided when a patient was a candidate for the 2-dialysis modalities.  3.Absence of information: no information about dialysis modalities was provided. Domain 3: Patients’ perception of the shared decision-making process of the dialysis modality  1. Shared decision: patient perception of choice made voluntarily after receiving complete information and advice about dialysis modalities.  2. Lack of shared decision: a patient who received complete information but had the perception of not participating in the decision-making process about the dialysis modality. *Bibliographic references.[15–19] HD = hemodialysis, PD = peritoneal dialysis. 2 Salas-Gama et al. • Medicine (2022) 101:42www.md-journal.com modalities was provided. The third domain was related to Table 2 summarizes the participants’ main characteristics, patients’ perceptions of the shared decision-making process distributed according to the dialysis modality. Most partici- of the dialysis modality. We defined a shared decision as the pants were male (71% PD, 63% HD). The median age in the patient’s perception of a voluntary decision after receiving com- PD group was lower than that in the HD group (68 years, inter- plete information and advice, and a lack of shared decision as quartile range 56–72 vs 72 years, interquartile range 60–83, a patient who received complete information but had the per- respectively), although this difference was not statistically sig- ception that they did not actively participate in the decision nificant. The percentage of patients who were not candidates for process. KT was higher among those treated with HD (56% HD, 21% To collect data related to the second and third domains, we PD, P = .003). No other differences between the 2 groups were designed a semi-structured questionnaire to assess the appropri- statistically significant. ate use according to the patient’s perspective (see Supplemental Dialysis modality selection was appropriate in 110 (78%, Digital Content 1, http://links.lww.com/MD/H564). It was 95% confidence interval [CI]: 69%–85%) patients and poten- pilot-tested before starting the study with 6 patients who began tially inappropriate in 31 (22%, 95% CI: 15%–31%) partici- dialysis. The feedback from these patients allowed us to evaluate pants (Fig. 1). Appropriateness was 96% for PD and 74% for their comprehension of the questions and make corrections to HD. The reasons for the appropriate dialysis modality selection improve their understanding. The questionnaire was repeated were as follows: patients ending up on the only possible modality with the same 6 patients 6 weeks later to assess the consis- because of contraindications for the other option (40 patients, tency and reproducibility of their answers, obtaining a kappa 28%) and the voluntary decision of a dialysis modality made by index >0.6, indicating good concordance. a fully informed patient (70 patients, 50%) (Table 3). Among We assessed the second domain related to the degree of infor- the patients with contraindications for PD, 10 (7%) were unable mation with items 5, 6, and 7 of the questionnaire, and the to perform PD at home. This inability was due to barriers to third domain related to voluntary choice and decision process self-care in 3 cases and home characteristics in 7 cases. Among with items 8 to 15, and 19 (see Supplemental Digital Content the 70 (50%) patients suitable for PD who voluntarily chose a 1, http://links.lww.com/MD/H564). We administered the ques- modality of dialysis after obtaining complete information, only tionnaire face-to-face to patients living in the city. Patients living one-third (16%) chose PD. outside the city area were invited to participate by their usual The reasons for the 31 participants with a potentially inap- healthcare team. After signing the informed consent form, they propriate dialysis modality selection are presented in Table 3. completed the questionnaire by telephone. The most frequent reason was incomplete information or omis- We defined appropriate dialysis modality selection when sion of information about both dialysis modalities, followed by patients ended up on the proper modality by voluntarily choos- the perception of a lack of shared decision-making. In the group ing one option after receiving complete information about either of 24 patients who started treatment with PD, only 1 patient modality or ending up on the only viable option because of con- met the criteria for inappropriate selection, which was due to traindications to the other modality. We considered potentially the perception of a lack of shared decision-making. Therefore, inappropriate dialysis modality selection in the following sit- inappropriate selection of dialysis modality was predominantly uations: incomplete or no information about the 2 modalities observed in patients treated with HD. If these patients would of dialysis; complete information given but patients’ perception have appropriately selected the modality, the number of patients of lack of shared decision-making; and non-prescription of the with PD could have doubled the current one (24 patients, 17% modality chosen voluntarily by the patient. vs 54 patients, 38%). The statistical analysis was mainly descriptive, with frequen- Table 4 shows the associations between patient characteris- cies and proportions for qualitative variables and central ten- tics and the appropriateness of the process of dialysis modality dency and dispersion measures for quantitative variables. We selection. We found that age, starting with HD, and not being a used basic bivariate tests to determine the association between candidate for KT were risk factors for potentially ending up on the appropriate dialysis modality selection and patient charac- dialysis’s inappropriate modality. In contrast, follow-up in the teristics. The association was further analyzed by multivariate ESKD consultation and information on RRT modalities were logistic regression using the backward stepwise procedure to associated with ending up on appropriate dialysis. Multivariate select predictive variables of appropriateness. The significance analysis confirmed that age was an independent predictor level for the inclusion of variables was 5%, and we used STATA of potentially inappropriate selection of dialysis modalities V.14.0. software for statistical analysis. (adjusted odds ratio 1.05, 95% CI 1.01–1.08, P = .011), and The Clinical Research Ethics Committee of the Nephrology information on dialysis options was a protective factor against Department approved the study, and we followed the potentially inappropriate selection (adjusted odds ratio 0.20, Strengthening the Reporting of Observational Studies in 95% CI 0.08–0.48, P < .001). Epidemiology statement.[32] 4. Discussion 3. Results This study found that the selection of dialysis modalities was During the 2 years of the study, 237 patients started RRT with appropriate in 78% of the study participants, and inappropriate either HD or PD. We excluded patients who died at the time in the remaining 22%, mainly among patients treated with HD. of the study, those who were unable to contact (most of whom To our knowledge, this is the first study to explicitly quantify lived outside the city of Barcelona), and those whose clinical the magnitude and causes of the appropriate and inappropriate situation prevented them from answering the questionnaire selection of dialysis modalities. (see Supplemental Digital Content 3, http://links.lww.com/ The proportion of contraindications for PD was in 17% to MD/H566, which shows the flowchart of patients). The study 28% range described in the literature,[12,19] and the most fre- group consisted of 141 patients: 24 (17%) who underwent PD quent contraindication in our study was abdominal-peritoneal and 117 (83%) who underwent HD at day 90. The excluded pathology. This result is consistent with that of Oliver et al.[33] patients did not differ significantly from the included patients in However, Jager et al noted that the most frequent contraindica- terms of age, sex, educational level, functional status, comorbid- tion for PD was the inability to perform PD independently.[27] ities, etiology of kidney disease, CKD presentation, information In contrast, we observed that only 2% of the patients had self- on RRT modalities provided, dialysis modality, and inclusion on care barriers and could not perform PD at home. Forbes et al[34] the waiting list for kidney transplant. found that lack of adequate housing was a significant barrier to 3 Salas-Gama et al. • Medicine (2022) 101:42Medicine Table 2 Characteristics of patients starting dialysis by dialysis modality. PD HD n = 24 (17%) n = 117 (83%) n % n % P value Sex .479  Male 17 71 74 63  Female 7 29 43 37 Age, median (IQR) 68 (56 - 72) 72 (60 - 83) .096 Distance to the nephrology department .178  <5 km 11 46 75 64  Five a 20 km 9 37 25 21  More than 20 km 4 17 17 15 Living with .108  Alone 2 9 10 9  With partner 20 83 79 67  With other companions 1 4 26 22  In a care facility 1 4 2 2 Educational level .147  Illiterate/Primary education 7 29 36 31  Secondary education 2 8 29 25  Middle or higher education 15 63 52 44 Karnofsky scale .603 ≥60 24 100 110 94 <60 0 0 7 6 Chronic kidney disease etiology .373  Diabetic nephropathy 5 21 16 14  Glomerular disease 5 21 14 12  Polycystic kidney disease 0 0 9 8  Vascular 3 12 10 9  Other 4 17 17 14  Unknown etiology 7 29 51 43 Comorbidities .108  None 10 42 26 22  One comorbidity 8 33 40 34  Two or more comorbidities 6 25 51 44 Kidney transplant .003  Previously transplanted or KT candidate 19 79 52 44  Not candidate 5 21 65 56 HD = hemodialysis, IQR = interquartile range, KT = kidney transplant, PD = peritoneal dialysis. Haemodialysis 74% (n=87) 26% (n=30) 4% Peritoneal dialysis 96% (n=23) (n=1) Total 78% (n=110) 22% (n=31) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Appropriate selecon Potenally inappropriate selecon Figure 1. Degree of appropriate and potentially inappropriate selection of dialysis modalities. implementing PD, reporting that only 29% of homes had the department, no home visits are conducted to assess whether the necessary characteristics for this technique. Only 5% of the par- domicile is suitable for PD. ticipants in our study reported home characteristics that contra- In our study, among patients with appropriate selection of indicated PD use. Unlike other centers,[25,35] in our nephrology dialysis modality without contraindications, despite receiving 4 Salas-Gama et al. • Medicine (2022) 101:42www.md-journal.com Table 3 Reasons for appropriate and potentially inappropriate selection of dialysis modalities. Dimensions Reasons n % Patients ending up on the 1. Patients with contraindications for dialysis Abdominal-peritoneal pathology 21 14.9% appropriate modality modalities: of dialysis  - Contraindications for HD (n = 0) Severe relative contraindications for PD 9 6.4%  - Contraindications for PD (n = 40/141; 28%) Unable to perform PD at home, due to: 10 7.1% a)Self-care problems (n = 3) b)Home characteristics (n = 7) 2. Patients without contraindications or barriers, Voluntarily choose HD 47 33.3% with complete information and with a voluntary Voluntarily choose PD 23 16.3% decision (n = 70/141; 50%) Patients ending up on the 3. Patients without contraindications or barriers, Incomplete information about the 12 8.5% potentially inappropri- with deficiencies in the information or the shared possible modalities (Started modality: ate modality of dialysis decision-making process (n = 31/141; 22%) HD = 12; PD = 0) Absence of information about the 9 6.4% possible modalities (Started modality: HD = 9; PD = 0) Complete information but the perception 10 7.1% of lack of shared decision-making (Started modality: HD = 9; PD = 1) Total 141 100% HD = hemodialysis, PD = peritoneal dialysis. complete information about the 2 dialysis modalities and after proposed a 6-step quality improvement initiative to maximize a shared decision-making process, 2-thirds voluntarily opted for PD utilization.[39] HD. The remaining third of the patients without contraindica- Our study showed an independent association between age tions and who were correctly informed voluntarily opted for PD. and potentially ending up on inappropriate dialysis. Previous It would be interesting to investigate why the patients volun- studies have described the underuse of PD in older adults.[26,27] tarily rejected DP. According to expert opinion,[12,13] PD should The authors reported that the principal reasons for PD underuse reach 30% to 45% of the proportion of patients starting RRT in this age group were non-clinical and consisted mainly of and even 50% if an adequate decision-making process is carried incomplete information regarding dialysis modalities and bar- out.[24,25] Therefore, our results also suggest that higher percent- riers to PD. In our study, correctly informed patients without ages of PD can be achieved using well-designed strategies based contraindications for PD, but with a perception of a lack of on patient information and a more assertive decision-making shared decision-making had a mean age of 81. Although age is process, always respecting patients’ values and preferences. a non-modifiable characteristic, it is not a contraindication for Among patients with inappropriate selection of dialysis the use of PD; therefore, information on both modalities should modality modalities, almost all started HD. When analyzing be offered indistinctly to older adults.[24] Another independent the drivers of potential inappropriateness, we observed that the association found in our study regarding the potential underuse most frequent reason was not receiving adequate information of PD was information about dialysis modalities. This variable about dialysis modalities. A study conducted by the Spanish was a strong protective factor against inappropriate selection Nephrology Nursing Society[36] reported that 65% of renal and highlighted the importance of an ESKD consultation unit patients perceived HD as well (or very well), but only 20% per- for increasing PD use. This finding agrees with the efficacy of ceived to know PD well (or very well). In our study, the second patient-targeted educational interventions in promoting PD, as reason for patients ending up on a potentially inappropriate described in the literature.[37] Since none of our objectives was dialysis modality, and closely related to the level of information, to analyze the causes of inappropriate modality selection, it is was the perception of a lack of shared decision-making. This possible that we may have missed other potential reasons that patient’s perception of unsatisfactory participation in choos- could explain the issue. ing the type of dialysis indicates the need to improve the deci- In the literature, there are nuances and variability in defin- sion-making process regarding the dialysis modality. In some ing clinical situations as contraindications to or barriers to PD. nephrology departments, these limitations can be explained It would help gather more scientific evidence and consensus to by structural, resource, or shortcomings. Simultaneously, some specify rational criteria for PD contraindications and barriers, physicians may practice a more conservative PD use, especially and reduce heterogeneity. Additionally, it would be interesting to in patients with advanced age or with specific clinical character- qualitatively assess the reasons, values, and preferences of patients istics. Some authors have described this as physician bias and with ESKD who voluntarily choose HD over PD when they can suggested that it could be a significant modifiable factor pro- choose between the 2 dialysis modalities. As future implications moting PD rates.[29] for clinical practice, it is essential to promote ESKD consultations To increase the appropriate use of dialysis modalities, the that follow a structured process that provides adequate infor- process of decision-making should be through integrated mation on dialysis modalities and is supported by decision aids interdisciplinary renal care programs and interventions.[29,37] to facilitate a shared decision-making process. Integrating home Several decision aids have been designed to facilitate a shared assessments for possible PD candidates would also help estimate decision-making process.[38] A study by Prieto-Velasco et al a high percentage of home problems reported by patients. conducted in 26 Spanish hospitals found that after imple- This study is the first to explicitly quantify the appropriate and menting a structured shared decision-making program, the potentially inappropriate use selection of dialysis modalities. We choice between the 2 dialysis modalities led to a 50/50 dis- conducted a comprehensive analysis of clinical records and com- tribution between PD and HD.[25] Mann et al recently pub- plemented this information using a patient questionnaire. This lished a scoping review of strategies to increase PD use and approach resulted in a valid and reliable tool for evaluating the 5 Salas-Gama et al. • Medicine (2022) 101:42Medicine Table 4 Association between patients’ characteristics and the potentially inappropriate selection of dialysis modalities. Appropriate Potentially inappropriate n = 110 (78%) n = 31 (22%) n % n % OR CI 95% P value Sex 0.29–1.64 .398  Male 69 76 22 24 1  Female 41 82 9 18 0.69 Age, median (IQR) 69 (58–78) 78 (62–85) 1.03 1.01–1.07 .04 Distance to the hospital  <5 km 67 78 19 22 1  Five to 20 km 28 82 6 18 0.76 0.27–2.09 .59  More than 20 km 15 71 6 29 1.41 0.48–4.13 .531 Living with  Alone 7 58 5 42 1  With partner 81 82 18 18 0.31 0.09–1.10 .068  With other companions 20 74 7 26 0.49 0.12–2.06 .33  In a care facility 2 67 1 33 0.7 0.05–10.01 .793 Educational level  Illiterate/Primary education 54 81 13 19 1  Secondary education 25 81 6 19 0.99 0.34–2.93 .996  Middle or higher education 31 72 12 28 1.61 0.65–3.96 .301 Karnofsky scale grouped ≥ 60 104 78 30 22 1 < 60 6 86 1 14 0.58 Chronic kidney disease etiology 0.07–4.99 .618  Diabetic nephropathy 18 86 3 14 1 0.10–4.75 .721  Glomerular disease 17 89 2 11 0.71  Polycystic kidney disease 7 78 2 22 1.71 0.23–12.55 .596  Vascular 9 69 4 31 2.67 0.49–14.56 .257  Other 17 81 4 19 1.41 0.27–7.26 .68  Unknown etiology 42 72 16 28 2.29 0.59–8.83 .23 Comorbidities  None 32 89 4 11 1  One comorbidity 32 67 16 33 4 1.20–13.28 .024  Two or more comorbidities 46 81 11 19 1.91 0.56–6.54 .301 Chronic kidney disease presentation 0.28–1.46 .296  Planned 35 73 13 27 1  Unplanned 75 81 18 19 0.65 End-stage kidney disease consultation 0.14–0.76 .009  No visits 26 63 15 37 1  Periodic visits 84 84 16 16 0.33 Information of RRT modalities provided 0.11–0.59 .001  Not recorded on the clinical chart 26 60 17 40 1  Recorded on the clinical chart 84 86 14 14 0.25 Starting dialysis modality 1.03–61.28 .047  Peritoneal dialysis 23 96 1 4 1  Hemodialysis 87 74 30 26 7.93 Kidney transplant 1.33–7.48 .009  Previously transplanted or KT candidate 62 87 9 13 1  Not a candidate 48 69 22 31 3.16 CI = confidence interval, IQR = interquartile range, KT = kidney transplant, OR = Odds ratio, RRT = renal replacement treatment. appropriateness of dialysis modality selection, incorporating the 5. Conclusions patient’s perspective and evidence recommendations. However, our In conclusion, our study on the appropriateness of the dialysis approach has some limitations. First, the study was retrospective in modality selection process for ESKD has the potential to increase nature, as we relied on information entered in the clinical records. the use of the PD modality. With better education and shared Although we supplemented this information with data from the decision-making, the number of patients on PD in our study patients’ questionnaires, patient recall when they started dialy- could potentially have doubled the current one. Interdisciplinary sis may be incomplete or inaccurate. The number of participants consultations with structured information processes and shared included in the study was the one estimated with the sample cal- decision-making should be promoted. Analysis of appropriate- culation. However, given the small number of patients in the study, ness is a good tool for studying PD selection patterns and iden- statistical interpretation and generalizability of the results should tifying opportunities to extend its use. be considered with caution. Our findings are also center-specific, given the variation in how education and decision-making are undertaken. A multicenter study with more participants and dif- ferent approaches to education and decision-making in other set- Acknowledgments tings would support these associations, ensure the usability of our We wish to thank the nephrology teams at the hospitals and approach, and reinforce the generalizability of our results. hemodialysis centers that collaborated in recruiting participants. 6 Salas-Gama et al. • Medicine (2022) 101:42www.md-journal.com Author contributions [17] Golper TA, Churchill D, Blake P, et al. NKF-K/DOQI clinical practice guidelines for peritoneal dialysis adequacy: update 2000. Am J Kidney Conceptualization: Karla Salas-Gama, Juan-Manuel Díaz- Dis. 2001;37(1 Suppl 1):S65–S136. Gómez, Ignasi Bolíbar Ribas. [18] NICE. Chronic kidney disease (stage 5) Clinical Guideline: perito- Data curation: Karla Salas-Gama, Juan-Manuel Díaz-Gómez. neal dialysis [Internet]. 2011 Available from: https://www.nice.org.uk/ Formal analysis: Karla Salas-Gama. guidance/cg125. Investigation: Karla Salas-Gama. [19] Blake PG, Quinn RR, Oliver MJ. Peritoneal dialysis and the process of Methodology: Karla Salas-Gama, Juan-Manuel Díaz-Gómez, modality selection. Perit Dial Int. 2013;33:233–41. Ignasi Bolíbar Ribas. [20] National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am Project administration: Karla Salas-Gama. J Kidney Dis. 2002;39(2 Suppl 1):S1–S266. Supervision: Ignasi Bolíbar Ribas. [21] Stanley M, Kelly J, Harris D. Acceptance onto dialysis - CARI guide- Validation: Ignasi Bolíbar Ribas. lines. Aust Fam Physician. 2007;36:439. Writing – original draft: Karla Salas-Gama. [22] Canadian Medical Association. Appropriateness in Health Care. CMA Writing – review & editing: Juan-Manuel Díaz-Gómez, Ignasi Policy. 2015. Available at: https://policybase.cma.ca/viewer?file=%2F- Bolíbar Ribas. media%2FPolicyPDF%2FPD15-05.pdf#page=1. [Access date October 6, 2021] [23] Cooper Z, Sayal P, Abbett SK, et al. A conceptual framework for appropriateness in surgical care: reviewing past approaches and look- References ing ahead to patient-centered shared decision making. Anesthesiology. [1] Bikbov B, Purcell CA, Levey AS, et al. Global, regional, and 2015;123:1450–4. national burden of chronic kidney disease, 1990–2017: a system- [24] Woodrow G. What are the factors underlying the variation in the use of atic analysis for the Global Burden of Disease Study 2017. Lancet. peritoneal dialysis? Nephrol Dial Transplant. 2013;28:501–4. 2020;395:709–33. [25] Prieto-Velasco M, Quiros P, Remon C, et al. The concordance between [2] Karopadi AN, Mason G, Rettore E, et al. Cost of peritoneal dialy- patients’ renal replacement therapy choice and definitive modality: is it sis and haemodialysis across the world. Nephrol Dial Transplant. a Utopia? PLoS One. 2015;10:e0138811. 2013;28:2553–69. [26] Tesar V. Peritoneal dialysis in the elderly--is its underutilisation justi- [3] Klarenbach SW, Tonelli M, Chui B, et al. Economic evaluation of dialy- fied? Nephrol Dial Transplant. 2010;25:3473–6. sis therapies. Nat Rev Nephrol. 2014;10:644–52. [27] Jager KJ, Korevaar JC, Dekker FW, et al. The effect of contraindications [4] Lameire N, Van Biesen W, Vanholder R. The role of peritoneal dialysis and patient preference on dialysis modality selection in ESRD patients as first modality in an integrative approach to patients with end-stage in The Netherlands. Am J Kidney Dis. 2004;43:891–9. renal disease. Perit Dial Int. 2000;20(Suppl 2):S134–41. [28] van Biesen W, Veys N, Lameire N, et al. Why less success of the peri- [5] Locatelli F, Marcelli D, Conte F, et al. Survival and development of car- toneal dialysis programmes in Europe? Nephrol Dial Transplant. diovascular disease by modality of treatment in patients with end-stage 2008;23:1478–81. renal disease. J Am Soc Nephrol. 2001;12:2411–7. [29] Hingwala J, Diamond J, Tangri N, et al. Underutilisation of peritoneal [6] Mazzuchi N, Fernández-Cean JM, Carbonell E. Criteria for selection of dialysis: the role of the nephrologist’s referral pattern. Nephrol Dial ESRD treatment modalities. Kidney Int. 2000;57(Suppl 74):S136–43. Transplant. 2013;28:732–40. [7] Vale L, Cody J, Wallace S, et al. Continuous ambulatory peritoneal dial- [30] Lameire N, Van Biesen W. Epidemiology of peritoneal dialysis: a story ysis (CAPD) versus hospital or home haemodialysis for end-stage renal of believers and nonbelievers. Nat Rev Nephrol. 2010;6:75–82. disease in adults. Cochrane Database Syst Rev. 2004;2004:CD003963. [31] Goodkin DA, Bragg-Gresham JL, Koenig KG, et al. Association of [8] Yeates K, Zhu N, Vonesh E, et al. Hemodialysis and peritoneal dialysis comorbid conditions and mortality in hemodialysis patients in Europe, are associated with similar outcomes for end-stage renal disease treat- Japan, and the United States: the Dialysis Outcomes and Practice ment in Canada. Nephrol Dial Transplant. 2012;27:3568–75. Patterns Study (DOPPS). J Am Soc Nephrol. 2003;14:3270–7. [9] Korevaar JC, Feith GW, Dekker FW, et al. Effect of starting with hemo- [32] von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting dialysis compared with peritoneal dialysis in patients new on dialysis of Observational Studies in Epidemiology (STROBE) statement: guide- treatment: a randomised controlled trial. Kidney Int. 2003;64:2222–8. lines for reporting observational studies. Int J Surg. 2014;12:1495–9. [10] Lorenzo V, Perestelo L, Barroso M, et al. Evaluación económica [33] Oliver MJ, Garg AX, Blake PG, et al. Impact of contraindications, bar- de la hemodiálisis. Análisis de los componentes del coste basado en riers to self-care and support on incident peritoneal dialysis utilisation. datos individuales [Economicevaluation of haemodialysis. Analysis Nephrol Dial Transplant. 2010;25:2737–44. of cost components based on patient-specific data]. Nefrologia. [34] Forbes SH, McCafferty K, Lawson T, et al. Is lack of suitable housing a 2010;30:403–12. barrier to home-based dialysis therapy for patients with end-stage renal [11] Busink E, Kendzia D, Kircelli F, et al. A systematic review of the cost-ef- disease? A cohort study. BMJ Open. 2013;3:e002117. fectiveness of renal replacement therapies, and consequences for deci- [35] Prieto-Velasco M, IsnardBagnis C, Dean J, et al. Predialysis education sion-making in the end-stage renal disease treatment pathway. Eur J in practice: a questionnaire survey of centres with established pro- Health Econ. 2022;1:3. grammes. BMC Res Notes. 2014;7:730. [12] Bouvier N, Durand PY, Testa A, et al. Regional discrepancies in perito- [36] Celadilla O, Julve M, Vives A, et al. Evaluación de la información neal dialysis utilisation in France: the role of the nephrologist’s opinion recibida por el paciente que inicia diálisis no programada o procedente about peritoneal dialysis. Nephrol Dial Transplant. 2009;24:1293–7. de trasplante. XXXII Congreso SEDEN. 2007. Available at: https:// [13] Neil N, Walker DR, Sesso R, et al. Gaining efficiencies: resources and www.revistaseden.org/files/1745_7.pdf. [Access date October 6, 2021] demand for dialysis around the globe. Value Health. 2009;12:73–9. [37] Devoe DJ, Wong B, James MT, et al. Patient education and peritoneal [14] ERA-EDTA Registry: ERA-EDTA Registry Annual Report 2019. dialysis modality selection: a systematic review and meta-analysis. Am Amsterdam, the Netherlands: Amsterdam UMC, location AMC, J Kidney Dis. 2016;68:422–33. Department of Medical Informatics. 2021. Available at: https://www. [38] Hermosilla T, Quirós PL, Remón C, et al. Grupo ERC-Sevilla: Agencia era-online.org/registry/AnnRep2019.pdf. [Access date October 6, de Evaluación de Tecnologías Sanitarias de Andalucía. Herramienta 2021] de ayuda a la toma de decisiones de pacientes susceptibles de diáli- [15] Covic A, Bammens B, Lobbedez T, et al. Educating end-stage renal dis- sis. 2011. Available at: https://www.aetsa.org/download/publicaciones/ ease patients on dialysis modality selection: clinical advice from the antiguas/AETSA_2009-8_HATD_ERC.pdf. [Access date October 6, European Renal Best Practice (ERBP) Advisory Board. Nephrol Dial 2021] Transplant. 2010;25:1757–9. [39] Mann BS, Manns BJ, Barnieh L, et al. Peritoneal dialysis: a scop- [16] Viglino G, Neri L. Theory and reality in the selection of peritoneal dial- ing review of strategies to maximise PD utilisation. Perit Dial Int. ysis. Perit Dial Int. 2008;28:480–3. 2017;37:159–64. 7