Received: 10 February 2021 Revised: 8 July 2021 Accepted: 17 July 2021 DOI: 10.1111/1744-9987.13711 ORIGINAL ARTICLE Are continued policies of prioritizing native vascular access in patients on hemodialysis programs useful? Sara Ibañez Pallarès1,2 | Vicent Esteve Simo3 | Alina Velescu4 | Irati Tapia Gonzalez3,5 | Silvia Collado Nieto6 | Albert Clara Velasco4 1 Vascular Surgery Department, Hospital de Terrassa, Barcelona, Spain Abstract 2 Surgery and Morphologic Sciences The guidelines recommend establishing native vascular access as opposed to Department, Autonomous University of prosthetic or catheter-based access despite information relating to its effective- Barcelona (UAB), Barcelona, Spain ness being scarce from a patient-orientated perspective. We analyzed the effec- 3 Nephrology Department, Hospital de tiveness of a continued policy of native vascular access (CPNVA) in patients Terrassa, Barcelona, Spain 4 Vascular Surgery Department, Hospital undergoing hemodialysis. A retrospective, observational study, including del Mar, Barcelona, Spain 150 patients undergoing hemodialysis between 2006 and 2012 at our center, 5 Medicine Department, Autonomous and who underwent a CPNVA. Statistical analysis was based on treatment University of Barcelona, Barcelona, Spain intention. In 138 patients (92%), the first useful access (FUA) was native, and 6 Nephrology Department, Hospital del Mar, Barcelona, Spain in 12 patients (8%), it was prosthetic. In 50 patients (33.3%), more than one procedure had to be carried out in to order to achieve FUA. The probability of Correspondence dialysis occurring via a FUA was 67.1% and 45.3% at 1 and 5 years respectively. Sara Ibañez Pallarès, Vascular Surgery Department, Hospital de Terrassa, Over the follow-up period (mean time = 30 months), 84 patients (56%) Barcelona, Spain. required repairs or new access, extending the effectiveness of the CPNVA to Email: sara.ib.87@gmail.com 88.3% and 73.2% at 1 and 5 years respectively. The effectiveness of the CPNVA was reduced if the patient: required a catheter initially (HR: 3.6, p = 0.007); in cases of initially elevated glomerular filtration rate (HR: 1.1, p = 0.040); in cases of history of previous access failure before FUA (HR: 3.9, p = 0.001); and in female patients (HR: 2.4, p = 0.031). The long-term effectiveness of a CPNVA is high. However, the percentage of patients requiring diverse proce- dures in order to achieve FUA and the need for re-interventions yield the necessity to optimize preoperative evaluation and postoperative follow-up. KEYWORDS AV access, dialysis, follow-up, native vascular access, patient-orientated, survival 1 | INTRODUCTION mortality, particularly in patients who require a renal substitution treatment. One of the factors determining Chronic renal disease represents a public health issue the morbidity/mortality of patients undergoing hemodial- due to its high incidence rate, prevalence, and morbidity/ ysis is the vascular access that they have [1]. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2021 The Authors. Therapeutic Apheresis and Dialysis published by John Wiley & Sons Australia, Ltd on behalf of International Society for Apheresis, Japanese Society for Apheresis, and Japanese Society for Dialysis Therapy. Ther Apher Dial. 2021;1–7. wileyonlinelibrary.com/journal/tap 1 2  IBAÑEZ PALLARÈS ET AL. Establishing vascular access for useful and long-term access was indicated on a preferential basis in all hemodialysis has been a constant cause of concern for patients over another type of access (catheter, prosthe- nephrologists and vascular surgeons since the creation of sis, arteriovenous) based on a physical examination. As the first primary arteriovenous anastomosis in 1966, by a general rule, a fistula at wrist of the non-dominant Brescia [2]. Nowadays, arteriovenous fistula at wrist or at limb was indicated. Subsequently, where there is a need elbow still represents the initial access for many patients. for it, this would be carried out at elbow and, finally, on As opposed to native techniques, arteriovenous prosthe- the dominant limb following the same criteria. Informed ses started to be used for vascular access in the 1960s, consent was obtained from all patients in writing, with becoming the procedure of choice for many years in the approval by the Ethics Committee of the Hospital United States and Canada. This tendency began to see del Mar. correction from 1997, when the National Kidney Founda- Over the entire course of follow-up, and in the case of tion started to publish the KDOQI Guidelines [3]; the any complication, dysfunction or thrombosis of the objectives of the Guidelines included improvement of access impeding adequate dialysis, a new autologous hemodialysis patients management. technique was decided on, wherever possible, as a correc- One of the main recommendations of the KDOQI tive measure, otherwise as a new access. Over the study Guidelines is to establish native vascular access as the period, there was no protocol of the systemic vascular preferred method over other procedures (AV prosthesis, hemodynamic investigations in these patients. Finally, catheters), given its reduced complication rate and more there was an indication for implantation of an arteriove- prolonged patency. However, the majority of publications nous prosthesis or permanent catheter only in cases that underpin this recommendation refer to studies that where it was assumed that all possibilities for a native have not been carried out under an intention-to-treat pre- access had been exhausted. mise, and that frequently have as their sole objective to Vascular access considered to be useful for hemodial- describe the patency associated with specific surgical ysis was: patent; punctionable; free of complications; it techniques, more than their usefulness for hemodialysis allowed for effective hemodialysis in low rercirculation from a global perspective (patent access, which allows for conditions; it had correct urea clearance parameters; it adequate dialysis, without complications) [4–7]. As a had no pressure-related issues (arterial and venous). result, information on the real effectiveness of a contin- These usefulness criteria were taken into consideration ued policy of native vascular access (CPNVA) as proposed for both the first useful access (FUA) and in a cumulative by the KDOQI Guidelines is relatively scarce. form for CPNVA. In this respect, “end of follow-up with As such, the objective of our study was to evaluate effective CPNVA” was considered to be when the patient the long-term effectiveness of a CPNVA in our patient was able to undergo dialysis via an autologous access population undergoing a program of hemodialysis due to until death, transplant or study end date; whereas “end end-stage chronic renal disease. of follow-up with ineffective CPNVA” was considered to be when the patient required an arteriovenous prosthesis or permanent catheter to continue in the dialysis pro- 2 | P A T I E N T S A N D ME T H O D S gram. Initial failures of the CPNVA were considered to have occurred in patients who finally required initiation One hundred and fifty patients were included in the of hemodialysis via an arteriovenous prosthesis (AVP) study retrospectively, all undergoing hemodialysis due to due to it being impossible to create a FUA. end-stage chronic renal insufficiency at the Hospital del Statistical analysis was performed using the SPSS pro- Mar (Barcelona) between January 1, 2006 and December gram (Statistical Package for the Social Sciences, version 31, 2010. Four patients were excluded who had been 25, SPSS Inc). The effectiveness curves were calculated incorporated into the hemodialysis program, having vas- for hemodialysis of the FUA and the CPNVA in the form cular access that was established in other centers; one of survival curves. Previously, the predictive factors for patient was excluded who was indicated, from the start, reduced effectiveness of hemodialysis were analyzed for permanent vascular catheter insertion. using the Kaplan–Meier test and Cox regression, both for Referrals for patients to the Vascular Surgery Service the FUA and the CPNVA. All values of p < 0.05 were for assessment regarding dialysis access were carried out considered to be statistically significant. from the advanced chronic renal disease consultations, Of the 150 patients, 101 (67%) were male. The mean where patients presented with a deterioration in renal age at the start of dialysis was 62.9 (±14.9) years. 38.7% function with a mean glomerular filtration rate (as per were diabetic; 14.7% presented with a history of coronary MDRD-4) of less than 20 mL/min/1.73 m2, in accor- heart disease; and 39% with a history of cardiac insuffi- dance with international consensus. An initial native ciency. The mean baseline glomerular filtration rate at  IBAÑEZ PALLARÈS ET AL. 3 the point in time when the first vascular access (for 3 | RESULTS hemodialysis) was created was 10.2 ± 4.4 mL/min/1.73 m2. 52.7% of patients required a temporary or tunneled catheter A FUA was achieved in 138 patients (92%). In 50 patients in order to start hemodialysis before achieving an FUA. (33.3%), at least one access was required prior to achiev- The clinical characteristics of the study population are ing an FUA. The radiocephalic fistula (RCF) represented described in Table 1. the first autologous fistula carried out in 103 patients (68.7%), also being the FUA in 43.3% of cases. In 35 patients (31%), the FUA was a fistula at elbow. Finally, TABLE 1 Clinical characteristics of the study population in the case of 12 patients (8%) it was necessary, following Age (years)a 62.9 ± 14.9 repeated unsuccessful attempts to achieve native access, n (%) to implant an arteriovenous prosthesis in order to start renal substitution therapy. We achieve a mean follow-up Gender of 30 months (maximum follow-up: 105 months). Male 101 (67) The probability of a patients maintaining hemodialy- Female 49 (33.1) sis via the FUA, including the group of patients who High blood pressure 100 (66.7) required an initial arteriovenous prosthesis as immediate Diabetes mellitus 58 (38.7) failures, was 67.1% at 1 year, and 45.3% at 5 years Ischemic cardiopathy 22 (14.7) Cardiac failure 58(38.7) T A B L E 2 Predictors of failure of CPNVA with p value (p) and Cardiac arrhythmia 35 (23.3) hazard ratio (HR) with a 95% confidence interval Cerebrovascular disease 24 (16) p Cancer 15 (10) Associated factors to value Hazard ratio CPNVA survival (p) (confidence interval) First useful access (FUA) Native 138 (92) Female gender 0.031 2.4 (1.1–5.1) Arteriovenous prothesis 12 (8) Presence of previous 0.007 3.6 (1.4–8.9) temporal catheter Number of failed accesses before the FUA Failed native access 0.001 3.9 (1.7–8.5) prior to FUA None 102 (68) Initially hight 0.004 1.1 (1–1.2) One or more 48 (32) glomerular filtration Hemodialysis program started by temporal 79 (52.7) rate catheter Abbreviations: CPNVA, continuous policy of native vascular access; FUA, a Main age ± standard deviation. first useful access. F I G U R E 1 (a) Cumulative effectiveness of the FUA for hemodialysis. b) Cumulative effectiveness of CPNVA. Detailed, the percentages of effectiveness in the first 6 years of follow-up of the FUA and the CPNVA 4  IBAÑEZ PALLARÈS ET AL. F I G U R E 2 Cumulative effectiveness of a CPNVA for hemodialysis based on: (a) patient gender; (b) temporal catheter as the first initial access; (c) number of native accesses failed prior to FUA. Detailed in percentages the effectiveness at 5 years of follow-up. CPNVA, continuous policy of native vascular access; FUA, first useful access (Figure 1a). This probability was reduced significantly probability is 90% if there is no history of prior temporary among patients who had required a temporary catheter catheter, compared to 56% where this has been necessary in order to start hemodialysis (HR: 0.54, p = 0.018), and (Figure 2b); and 86.3% in patients who did not require in the case of patients with higher baseline glomerular any previous access, compared to 48% in those patients filtration rate values (HR: 1.8, p = 0.026). who required one or more prior accesses (Figure 2c). It is Over the course of follow-up, surgical repairs were noteworthy that a marginally significant association was carried out and/or new native accesses were created in found between reduced effectiveness and the presence of order to treat dysfunctions or losses of the FUA diabetes mellitus; however, no significant correlation has in 84 patients (56%); it was possible to increase the effec- been identified in the sample with respect to age or other tiveness of the CPNVA at 1 and 5 years by 88.3% and comorbidities. 73.2% respectively (Figure 1b). The main predictors of failure of CPNVA were: female sex (HR: 2.4, p = 0.031); history of temporary 4 | DISCUSSION catheter at initiation of the hemodialysis program (HR: 3.6, p = 0.007); increased baseline glomerular filtration The scientific literature on accesses for hemodialysis is rate values (HR: 1.1, p = 0.040); and, of course, pre- particularly prolific in articles relating to concrete surgi- senting with at least one failed access prior to the FUA cal or endovascular techniques, which are particularly (HR: 3.9, p = 0.001). These data are summarized in useful for evaluation of each of their specific contribu- Table 2. tions with respect to others, but which often do not allow Accordingly, at 5 years, the probability that a female for an overview. In contrast, natural history studies patient would maintain dialysis via a native access is allow for quantification and analysis, from a global and 59%, compared to 81% for a male patient (Figure 2a); the patient-orientated perspective, of the main problems  IBAÑEZ PALLARÈS ET AL. 5 associated with the durability and usefulness of accesses, rates of hospitalizations [14] and severe infections [15] in the probability of reintervention, or the possibility of hemodialysis population are described in patients with a eventually having a permanent catheter. Our study, in catheter. line with previous studies, contributes to a better under- In our group, 33% of patients required creation of standing of the real effectiveness of a CPNVA as set out more than one arteriovenous access in order to achieve a by the KDOQI Guidelines, and provides specific informa- first useful native access. In similar studies, around half tion on two aspects that have scarcely been the subject of of patients required more than on procedure in order to analysis to date: the number of interventions necessary to finally achieve a valid native access for dialysis [16, 17]. achieve useful access; and the durability/usefulness of a We were successful in creating a permanent first access preferential policy of autologous accesses over the long for hemodialysis as an arteriovenous fistula in 92%, a per- term, carrying out an analysis based on treatment centage much higher than the target of 65% set by the intention. Fistula First Initiative [16, 18]. Despite this, 8% of our With respect to the characteristics of the study popu- patients presented as initial failures under this policy, lation, age on initiation of dialysis, and sex distribution, requiring an arteriovenous prosthesis to be inserted. This there were no great differences as compared with these fact contrasts with the prevalence of dialysis via pros- aspects in studies in the Mediterranean region [8]. In pre- thetic access in the United States at around 20%, vious studies in the United States, a substantially lower according to data from 2013 [18] . age on initiation of dialysis could be identified (around At 1 year, 67% of the patients continued to receive 56 years old), as well as higher rates of comorbidities, hemodialysis via the FUA. Patel et al. and Biuckians such as ischemic cardiopathy [7]. In fact, increased sur- et al. describe similar rates of primary patency at 1 year vival has been described in patients on hemodialysis in [19, 20]. Thanks to successive treatment of dysfunctions the Mediterranean region as compared to other regions and thrombosis relating to the arteriovenous fistula by of Europe or the United States, which highlights the means of repair procedures or creation of new native importance of extending the maximum life of autologous access, we were able to achieve a probability of our vascular accesses [9, 10]. patients undergoing dialysis via a native access of 88.3% Following the recommendations from the aforemen- and 73.2% at 1 year and at 5 years respectively. To the tioned Guidelines, in our center the RCF represented the best of our knowledge, these data have not previously initial autologous access created in almost 69% of been described from this perspective and it seems, how- patients, with this type of fistula becoming the FUA in ever, to be information that should be highlighted, not almost half of these. The autologous access at elbow only for comparison with other therapeutic strategies, (including humerocephalic and humerobasilic) repre- but also to inform patients who are entering dialysis sents the first useful native access in almost 49%. In the programs. series described by Solesky et al. [11], the first access cre- With regard to factors that negatively influence sur- ated is radiocephalic in 20%. vival of native accesses, in the studies reviewed, age, In our population sample, the percentage of patients female sex, and presentation of diabetes are cited as asso- who initiated hemodialysis via a temporary catheter or ciated factors [12, 18]. In our study, the factors which tunneled catheter was 52.7%, a proportion which is simi- were associated with worse results in CPNVA in a statis- lar to data described in other studies and in the KDOQI tically significant way were: female sex; history of tempo- Guidelines [3, 4, 8, 10, 12]. It is well known that starting rary catheter to initiate dialysis; presenting with one or hemodialysis via a temporary catheter or tunneled cathe- more failed access prior to the FUA; and elevated glomer- ter is independently associated with worse results when ular filtration rate at the time when the access was cre- creating native vascular accesses. Ethier et al. describe ated. With regard to the factors of female sex and history the most important factor associated with initial catheter of previous catheter, the worsened progress may be use as the late referral to specialist, which varies attributed to poorer quality of the vascular network. according to the country [12]. Either et al. [12] have described an increased rate of fail- Nowadays, it seems to be in doubt that use of a cathe- ures of accesses with increased time delay between crea- ter by itself could increase the mortality rates in the tion and first cannulation (it can be understood that hemodialysis patients. Brown et al. [13] suggest that patients subject to procedures to create an access too the best baseline characteristics of the patients undergo- early have better clearing at the time of surgery), ing a native vascular access are related with a low mortal- although the causes associated with this failure are ity, compared to the group of patients in which we decide unknown. Other parameters were analyzed, such as the to implant a permanent catheter, whom usually are more presence of diabetes mellitus, age, hemoglobin levels at frail and elderly. However is widely known the highest the time when access was created, as well as other 6  IBAÑEZ PALLARÈS ET AL. comorbidities: no statistically significant differences were the elevated number of cases requiring a catheter to initi- identified in the survival curve for the accesses. ate dialysis, and the frequent need to repair or carry our One of the major limitations of this study is its retro- new autologous accesses over time bring to light the spective design, which has limited the number of preop- importance of optimizing clinical evaluation both preop- erative variables that it has been possible to include eratively and during follow-up. Detecting predictive fac- without loss of information or inclusion of poor quality tors of subsequent dysfunction may be of great benefit for data. Similarly, the small number of patients included in this purpose. New studies are necessary, however, to the analysis could have limited the detection of predictive evaluate whether adding complementary examinations factors to include only those that showed a close associa- (such as Doppler ultrasound scan) to the clinical exami- tion with later dysfunction of the access. This limitation, nation might contribute to improving continued policy of however, is common to the other studies that have evalu- native vascular access results. ated the results, similarly from a patient-orientated per- spective [11, 17]. CONFLICT OF INTEREST The ultrasound evaluation of the arteriovenous capi- The authors of this article have declared that there are no tal constitutes a very important tool in the choice of the potential conflicts of interest relating to the content of type of vascular access. In our case, the lack of this infor- this article. mation constitutes another limitation. Therefore, the aim of our study was to prioritize a follow-up as long as possi- ORCID ble, so we decided to include patients from 2006 to 2012, añez Pallarès https://orcid.org/0000-0003-2433- Sara Ib assuming that on these periods the ultrasound preopera- 4876 tive evaluation was not performed by routine in our cen- Vicent Esteve Simo https://orcid.org/0000-0003-1842- ter. We are analyzing a more recent cohort in which the 3442 ultrasound evaluation could have a decisive role in the indication of the vascular access. RE FER EN CES Moreover, prosthetic access was also considered as an 1. Anton G, Pérez P, Alonso F, Nicanor V. Accesos vasculares en initial failure of our patient—focused program and was hemodialisis: un reto por conseguir. Nefrologia. 2012;32:103–7. not excluded from the initial analysis in despite that their  2. Gonzalez Alvarez MT, Martínez Cerc os R. Manual de accesos results could be, a priori, more favorable. In our opinion, vasculares para hemodialisis. Barcelona: Marge Books; 2010. their inclusion showed a true global vision and real bur- 3. KDOQI Clinical practice guidelines for vascular access. Update den of the CPNVA strategies in our institution. However, 2006. National Kidney Foundation. http://www2.kidney.org/ the inclusion of these prosthetic accesses could be inter- professionals/KDOQI/guideline_upHD_PD/VA. preted as another limitation or bias selection and could 4. Sidawy AN, Spergel LM, Besarab A, Allon M, Jennings WC, detect some differences in the predictive factors involved. Padberg FT Jr, et al. The Society for Vascular Surgery: clinical practice guidelines for the surgical placement and maintenance Finally, we have data available to us on the incidence of arteriovenous hemodialysis access. J Vasc Surg. 2008;48 of a catheter on initiation of dialysis, but not on the prev- (Suppl):2S–5S. alence of dialysis via catheter in our study population: 5. Lok CE, Sontrop JM, Tomlinson J, Rajan D, Calttral M, this information would be interesting to obtain as a qual- Oreopoulos J, et al. Cumulative patency of contemporary fistu- ity measure relating to overall care to the dialysis patient. las versus grafts (2000–2010). Clin J Am Soc Nephrol. 2013;8: Similarly, only those patients who were evaluated by the 810–8. vascular surgeon have been included in the study, and as 6. Murad MH, Elamin MB, Sidawy AN, Malaga G, Rizvi AZ, such, it does not take account of the small proportion of Flynn DN, et al. Autogenous versus prosthetic vascular access the patient population which, due to their overall condi- for hemodialysis: a systematic review and meta-analysis. J Vasc Surg. 2008;48:34S–47S. tion, are maintained on a dialysis regime via catheter 7. Dixon B, Novak S, Fangman J. Hemodialysis vascular access with no view to creating an autologous or prosthetic vas- survival: upperarm native arteriovenous fistula. Am J Kidney cular access. Dis. 2002;39:92–101. 8. Pantelias K, Grapsa E. Vascular access today. World J Nephrol. 2012;1:69–78. 5 | C ON C L U S I ON S 9. Rodriguez JA, Ferrer E, Olmos A, Codina S, Borrellas J, Piera L. Analysis of the survival of permanent vascular access. From a patient-focused perspective, the long-term effec- Nefrología. 2001;21:260–73. tiveness of a continued policy of native vascular access is 10. Rodríguez JA, Ardamans L, Ferrer E, Olmos A, Codina S, high. However, the proportion of patients requiring more Bartolomé J, et al. The function of permanent vascular access. than one access before achieving an first useful access, Nephrol Dial Transplant. 2000;15:402–8.  IBAÑEZ PALLARÈS ET AL. 7 11. García MJ, Viedma G, Sanchez MC, Borrego FJ, Borrego J, 17. Nguyen TH, Bui TD, Gordon IL, Wilson SE. Functional Perez del Barrio E, et al. Permanent vascular access in elderly patency of autogenous AV fistulas for hemodialysis. J Vasc patient who starts on hemodialysis: fistulae or catheter? Access. 2007;8:275–80. Nefrologia. 2005;25:307–14. 18. Jennings WC. Creating arteriovenous fistulas in 132 consecutive 12. Radoui A, Lyoussfi Z, Haddiya I, Skalli Z, El Idrissi R, patients: exploiting the proximal radial artery arteriovenous fis- Rhou H, et al. Survival of the first arteriovenous fistula in tula: reliable, safe, and simple forearm and upperarm hemodi- 96 patients on chronic hemodialysis. Ann Vasc Surg. 2011; alysis. Arch Surg. 2006;141:27–32. 25:630–3. 19. Papanikolaou V, Papagiannis A, Vrochides D, Imvrios G, Gakis D, 13. Brown RS, Patibandla BK, Goldfarb-Rumyantzev AS. The Fouzas I, et al. The natural history of vascular access for hemodial- survival benefit of “fistula first, catheter last” in hemodi- ysis: a single center study of patients. Surgery. 2009;145:272–9. alysis is primarily due to patient factors. J Am Soc 20. Van de Luijtgaarden MVM, Noordzij M, Stel VS, Ravani P, Nephrol. 2017;28:645–52. https://doi.org/10.1681/ASN. Jarraya F, Collart F, et al. Effects of comorbid and demographic 2016010019 factors on dialysis modality and related patient survival in 14. Lacson E Jr, Wang W, Lazarus JM, Hakim RM. Change in vas- Europe. Nephrol Dial Transplant. 2011;26:2940–7. cular access and hospitalization risk in long-term hemodialysis patients. Clin J Am Soc Nephrol. 2010;5:1996–2003. https://doi. org/10.2215/CJN.08961209 15. Sahli F, Feidjel R, Laalaoui R. Hemodialysis catheter-related How to cite this article: Ibañez Pallarès S, Esteve infection: rates, risk factors and pathogens. J Infect Simo V, Velescu A, Tapia Gonzalez I, Collado Public Health. 2017;10:403–8. https://doi.org/10.1016/j.jiph. Nieto S, Clara Velasco A. Are continued policies of 2016.06.008 prioritizing native vascular access in patients on 16. Flu H, Breslau PJ, Krol-van Straaten JM, Hamming JF, hemodialysis programs useful? Ther Apher Dial. Lardenoye JW. The effect of implementation of an optimized 2021;1–7. https://doi.org/10.1111/1744-9987.13711 care protocol on the outcome of arteriovenous hemodialysis access surgery. J Vasc Surg. 2008;48:659–68.