This is a retrospective, descriptive, observational and cross-sectional study of all prevalent patients on chronic haemodialysis from seven haemodialysis centres located outside hospitals in the regions of Madrid, Castille - León and Galicia conducted in June 2021. The study population included637 haemodialysis patients as of 30 June 2021. Patients were referred to the haemodialysis centre from the referral hospital. Both the creation of the first venous access and subsequent venous access depended on the surgery department of the referral hospital. If a patient arrived at the centre with a CVC, it was assessed and referred to the Vascular Surgery Department of the referral hospital, either directly or through the Nephrology Department.

All centres had a common protocol for VA monitoring and surveillance daily with first-generation methods and with second-generation methods quarterly using dilution screening methods to calculate periodic measurement of AVF blood flow (Qa) and Doppler ultrasound.

The following variables were collected: age, gender, nationality, dialysis centre, region, aetiology of chronic kidney disease (CKD), AVF and prosthesis thrombosis rate, the VA with which the patient started HD, the number of previous failed AVF, time since starting HD, time since first CVC placement, surgical status and the reasons for having a CVC. In patients whose reason was refusal to have an AVF, patients were asked directed questions about why they had refused.

For statistical analysis, Pearson's χ2 test was used for categorical variables, Student's t-test (parametric) or Mann-Whitney U-test (non-parametric) for continuous variables, or one-way ANOVA for multiple comparisons. Normality of distribution was verified by the Kolmogorov-Smirnov test. Continuous variables were expressed as mean (standard deviation) or median (interquartile range), as appropriate. Categorical variables were reported as numbers (percentage). Correlation was studied using Pearson's correlation coefficient. All p-values were two-tailed. Statistical significance was set at p < 0.05.

Of the 637 patients analyzed, 255 (40%) had a CVC, 346 (54.3%) had an AVF and 36 (5.7%) had a prosthesis. The mean age was 66.4 ± 15.4 years, 65% were male and 10% were of non-Spanish nationality. The aetiology of CKD was diabetic nephropathy in 27.5%, not recorded in 23.1%, glomerular in 15.2%, tubulointerstitial nephropathy in 8%, polycystic kidney in 5.7%, vascular in 11% and other in 9.4%.

Demographic characteristics and factors associated with CVC use in prevalent patients are shown in Table 1; Factors associated with CVC use are having started haemodialysis with a CVC (83.9% of prevalent patients with a CVC vs 40.8% of prevalent patients with an AVF/prosthesis [p < 0.001]). There were no significant differences in age, gender, nationality, or time on haemodialysis among prevalent CVC patients.

There were no significant differences between the Spain's regions (Table 1), but significant differences were found in relation to the referral hospital and centre (p = 0.004) (Fig. 1). The average autologous fistula thrombosis rate was 0.04 and the average prosthesis thrombosis rate was 0.22 across centres. No significant differences were found in the rate of thrombosis either by centre or by autonomous region (p = 0.076).

Fig. 1.

Percentage of prevalent patients and central venous catheter (CVC) incidents by centre (p = 0.004).

(0.15MB).

There was a positive correlation between starting HD with CVC and maintaining CVC thereafter (r: 0.41; p < 0.001).

Starting HD via CVC was significantly associated with an urgent initiation of HD and not having been followed and assessed in the advanced chronic kidney disease (ACKD) clinic (Table 2). No statistically significant differences were found in the use of CVC between patients who had been using them for more than six months (149/450; 33.1%) versus those using the CVC for less than six months (12/25; 48%); p = 0.303. Transplant recipients and those on the waiting list for living-donor transplant had a higher risk of starting HD with CVC.

Fig. 2 shows the flow of patients and the reasons why patients had CVC at the time of the study.

Fig. 2.

Reasons for patients having central venous catheters (CVC).

VA: vascular access; AVF: arteriovenous fistula.

• a)

  Pending surgery for other reasons, including inability to perform VA surgery due to hospitalisation, other surgeries, etc.

• b)

  Non-functioning VA, including: primary failure (9 patients - 3.5%) or recent thrombosis (2 patients - 0.78%).

• c)

  Non-puncturable AVF including: maturing (8 patients - 3.13%); non-puncturable brachiobasilic AVF pending superficialisation (5 patients - 1.96%); fracture of the AVF arm preventing its use (2 patients - 0.78%).

• d)

  Ruled out for vascular surgery, including: lack of vascular bed (25 patients - 9.8%), AVF closure due to ischaemia (4 patients - 1.56%), multiple previous thromboses (14 patients - 5.49%).

• e)

  Ruled out by nephrology, including: age and/or comorbidity (36 patients - 14.1%), rest from peritoneal dialysis (3 patients - 1.17%), possible recovery of renal function (1 patient - 0.39%), on waiting list for living-donor transplant (2 patients - 0.78%).

• f)

  Fear of AVF, including: fear of surgery and punctures due to previous bad experiences (31 patients - 12.1%) and fear of surgery and punctures without previous surgeries (27 patients - 10.5%).

• g)

  Preference for catheter, including: aesthetics, comfort (34 patient - 13.3%).

(0.5MB).

Table 3 shows the demographic characteristics of the different groups of prevalent patients with CVC.

Out of the 255 HD patients with CVC, 92 (36%) refused to have an AVF. Two groups of patients were identified: those who were afraid of having an AVF either because of the punctures or the surgery, and those who preferred CVC to AVF for aesthetic reasons or comfort. The most common reason for having fear was having experience previous complications. Table 4 shows the differences between patients who refused due to fear and those who refused because of preference. Patients preferring CVC were younger and 94% had started HD with CVC.

The main finding of this study is that, despite the recommendations of clinical practice guidelines,6,7 the percentage of CVC in HD units remains high (in our study 20% above the recommendation by guideline [<25%]), which is in line with other similar publications.1–4

The variability between countries and centres,2–4 as was also evident in this study, raises the following dilemma: are there modifiable factors that can be controlled to achieve the objectives proposed by the guidelines? Or by contrast, nowadays, with the new criteria of personalisation and the active incorporation of the patient in decision-making on VA,15 should we assume that it is not possible to achieve the proposed objectives? Could we reverse the situation by acting differently or improving processes? Or should we change the guidelines and bring them into close to the reality?

Different factors may be contributing to the high prevalence of CVC in units: some are dependent on the patient,16 others on the surgical team,17 others on structural factors, and even factors that have arisen in recent years, such as the influence of the COVID pandemic on the creation of VA.18 A reflection about these factors may help to guide us towards measures that can be implemented to improve VA outcomes by focussing on factors that are modifiable.

In this study, unavoidable patient-dependent factors such as age, gender or non-Spanish nationality did not influence the overall higher use of CVC in prevalent patients and at the start of HD, although the group of patients who were rejected for surgery were older than the rest. Age alone is not considered a contraindication for AVF, but it is a complex and highly-debated issue that requires a balance between patient preference, the likelihood of the VA working, and potential complications in the context of limited patient survival.19,20 Although this may highlight the real difficulty of performing an AVF in an increasingly ageing population, with more comorbidity and poor vascular health, it is also susceptible to subjectivity, both by the nephrologist and the vascular surgeon. In fact, the experience and involvement of the vascular surgeon not only influences VA outcomes, but also the decision to rule a out a patient for creation of AVF or even the patient's refusal to undergo an AVF, and it could be is a modifiable factor.21,22 Studies have shown that increased training and continuing education in VA for residents, nephrologists and vascular surgeons can prevent patient refusal of AVF.23–25

Organisation is a factor that can be modified by many of us. Different models of VA management and the relationship between vascular surgery and nephrology have been shown to have an enormous influence on VA outcomes.26 In this study, up to 30% of patients, despite being considered candidates for AVF/prostheses, were dialysed via a CVC. Two thirds of them were patients who were still waiting for surgery and were at different stages of the process (Fig. 1), which illustrate the need for an improvement on VA planning. In addition to glomerular filtration rate and its progression,6,26 other factors have to be considered when deciding when to refer a patient for placement of VA; for example, waiting times in the vascular surgery department or the characteristics of the patient and their environment.27 The decision needs to be made on an individual basis and patients should be referred to surgery early if they are at risk of slower maturation or having complications (females, advanced age, diabetes mellitus, peripheral arterial disease, smoking or obesity),28,29 or if they may require more complex techniques (superficialisation)30; also in those cases with comorbidities with a risk of rapid decrease in glomerular filtration rate that may require an urgent initiation of HD31 or if their social or family situation makes them to be at risk of being lost to the clinic follow-up.32

Another avoidable factor identified in the present study is the fact that the use of CVC at the start of HD is associated with the need of urgent start of HD with no previous follow-up in the advanced chronic kidney disease (ACKD) clinic.33 The early detection of CKD and the dissemination of knowledge in primary care and other specialities so that these patients are referred early to nephrology continues to be a major challenge.34

Of note is the high percentage of patients found in this study who refused to have an AVF created (36%), due to fearor preference as the most important reasons. This fear is associated with the high prevalence of anxiety and depression that affect dialysis patients and will influence their preferences.35 In a nationwide study that analysed the most common causes of concern about surgery, the most important causes were: fear of the unknown, possible complications, impact on quality of life and pain control.36 The scientific community considers AVF to be the optimal VA based on the lower rate of infection, morbidity and mortality rates and lower costs. However, patients give priority to how the VA may affect their quality of life and future prospects.37,38 The association between fear and number of previous AVF attempts suggests the need to reduce negative experiences for patients. Choosing the best VA and the expert hands of both surgeons and nurses are important. In those that refuse to accept an AVF, a significant association has been reported,and also found in the present study, between the preference for a CVC and the fact of either having a CVC or having had one at some point39,40 so another modifiable aspect to minimise refusals is to prevent patients from starting haemodialysis with a CVC. The choice of VA together with the choice of renal replacement therapy is a process that requires both information and time in ACKD clinics.41 We need to be able to identify the barriers that prevent this information from being provided in an appropriate and individualized manner to each patient, based on specific communication techniques.42

Lastly, adequate VA surveillance using first- and second-generation methods allows for early detection of stenosis and prevention of thrombosis, and is a modifiable factor that can influence the rate of CVC. In this study, the thrombosis rates of both AVF and prosthesis were low and below the values recommended by the guidelines of the Grupo Español Multidisciplinar del Acceso Vascular (GEMAV) [Spanish Multidisciplinary Group on Vascular Access].6

The limitations of this study include its retrospective nature and, since it deals with data collected on arrival at the peripheral centre, it is difficult to gather specific data relating to the procedure within the ACKD or transplant clinic (for example, glomerular filtration rate [GFR] at the time of referral, the person providing information to the patient and with whom the decision about VA is made). However, this study has the following strengths: it shows results from a large sample of patients, coming from different centres, hospitals and autonomous regions, which offers a much broader view than if they were a single centre and fosters reflection on future initiatives to improve VA.

In conclusion, there are still modifiable causes that influence the high percentage of patients with CVC, necessitating a systematic evaluation to identify barriers to the process of AVF creation. Strategies are needed to promote the planning, creation and maintenance of VA from the ACKD clinic, with the main objective that most patients can have HD with the best possible VA after a personalised assessment by the multidisciplinary team, to ensure that outcomes are not the result of variability of criteria or poor organisation, management or experience of the healthcare professionals responsible for the VA.