Patient perspective and satisfaction with different types of vascular access in hemodialysis: a systematic review and meta-analysis

Abstract

Background

To assess satisfaction and perspectives of adult hemodialysis (HD) patients on different vascular access (VA) types (AVF, AVG, CVC) by synthesizing quantitative studies using validated questionnaires, focusing on overall satisfaction, physical symptoms, social functioning, and complications.

Methods

We systematically searched Medline, Web of Science, and Scopus for studies assessing adult HD patients’ perspectives on VA using structured questionnaires. Risk of bias was assessed using the Newcastle-Ottawa Scale. Meta-analysis was performed using a random-effects model across three pairwise comparisons: AVF vs. CVC, AVF vs. AVG, and CVC vs. AVG.

Results

Of 1,253 records, 11 studies (n = 2910) met inclusion criteria. AVF patients had higher overall satisfaction compared to those with CVCs (MD: 7.06%, 95% CI: 2.98–11.14). Compared to AVG, AVF patients had higher overall satisfaction, though not statistically significant (MD: 5.12%, 95% CI: − 0.42 to 10.66; p = 0.069). Dissatisfaction with physical symptoms was significantly higher in AVF patients than in those with CVCs (MD: 8.76%, 95% CI: 4.19 to 13.34). Compared to CVC, AVF patients reported significantly lower dissatisfaction in both social functioning (MD: − 9.48%, 95% CI: − 15.98 to − 2.98) and dialysis-related complications (MD: − 10.08%, 95% CI: -12.66 to -7.50).

Conclusions

AVFs are associated with higher overall patient satisfaction compared to other VA types. While AVF patients experience more physical discomfort, this is outweighed by significantly greater satisfaction in areas such as social functioning and dialysis-related complications. These findings emphasize the importance of considering patient-reported outcomes when evaluating VA options in HD care.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12882-025-04393-9.

Background

Chronic kidney disease is a rapidly increasing global health problem with an estimated 9.1% prevalence and 35.8 million disability-adjusted life-years (DALYs) worldwide [1]. More than 2.5 million people currently receive kidney replacement therapy (KRT), a figure projected to rise to 5.4 million by 2030 [2]. The most common form of KRT worldwide is hemodialysis (HD), which accounts for up to 69% of all KRT and 89% of all dialysis treatments. To enable HD, patients require a form of vascular access (VA), including arteriovenous fistula (AVF), arteriovenous graft (AVG), or central venous catheter (CVC).

Chronic diseases are associated with a reduced quality of life, particularly in HD patients who spend significant time in dialysis units; therefore, assessment of satisfaction with care and VA is sine qua non. Various studies show that CVCs are linked with high mortality and morbidity, infection, dysfunction, and damage to central veins [3–5], but CVCs are still used frequently, especially by the incidence patients [6]. On the other hand, while acknowledging that AVFs offer the best outcomes for the overall HD population, providers are shifting away from a universal fistula-first approach and focusing on more patient-centered recommendations [7]. Furthermore, a recent study by Zhang et al. demonstrated that HD patients with lower VA satisfaction had a significantly higher all-cause mortality rate [8]. All of these factors have fueled a growing interest among physicians in understanding patients’ perspectives and satisfaction with various types of VA, an aspect that was previously only marginally addressed in earlier research [9]. Understanding patients’ perspectives on their VA is crucial for clinicians to address concerns, incorporate patient preferences into the decision-making process, ensure adequate knowledge transfer, and identify barriers to selecting the optimal VA.

Due to the lack of well-constructed questionnaires, a previous systematic review performed in 2014 could only address the qualitative studies, mainly with semi-structured interviews [10]. It showed that patients mostly describe their VA as a painful reminder of their body’s failures and the “abnormality” of being tethered to a machine, which disrupts their sense of identity and way of life. In recent years, with the development of valid and structured questionnaires regarding patient perspectives on the type of VA, it has become possible to conduct a systematic review and meta-analysis of quantitative studies.

This systematic review and meta-analysis aims to evaluate the perspectives and satisfaction of adult patients undergoing HD regarding different types of VAs, including AVF, AVG, and CVC. It synthesizes quantitative studies that use structured and validated questionnaires to identify differences in overall satisfaction and specific domains, such as physical symptoms, social functioning, and dialysis-related complications.

Methods

Literature search strategy

We performed a systematic search including English-language articles published up to April 8, 2025, in Medline, Web of Science, and Scopus databases. The search query was developed in consultation with a professional librarian and included terms such as “vascular access questionnaire”, “patient perspective”, “dialysis”, “arteriovenous fistula”, “central venous catheter”, and “arteriovenous graft”. To ensure comprehensive coverage, we additionally conducted a manual review of the reference lists from pertinent studies and review articles. No restrictions were applied regarding the date, publication status, or type of study. The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines and checklist were used in the preparation of this systematic review and meta-analysis [11].

Eligibility criteria

We included full-text, peer-reviewed original studies that used a well-defined and structured questionnaire specifically addressing the patients’ perspectives on the type of VA. All types of studies, including observational studies such as cross-sectional and cohort studies, were eligible for inclusion. Other inclusion criteria were studies with patients on HD with any VA and aged above 18. In addition to studies comparing major VA types (AVF, AVG, and CVC), we also included studies that examined variations within AVFs—such as differences in creation method (e.g., surgical vs. endovascular) and anatomical outcomes (e.g., aneurysmal vs. non-aneurysmal)—to capture a more comprehensive understanding of patient satisfaction and perspectives across the spectrum of VA experiences.

Conference abstracts, editorials, letters to the editor, narrative reviews, and case reports were excluded. Articles that used semi-structured questionnaires, generic quality-of-life assessments, and qualitative methodologies were also excluded. Studies that solely assessed satisfaction without conducting a comparative analysis of different types of VA were also excluded.

Study selection process

After removing duplicate records, two authors (AA and SP) independently reviewed the titles and abstracts of the articles using EndNote 21 software. Then, eligibility criteria were assessed through a review of the full text, and articles that did not appear to meet the criteria were excluded. Consultation with a third author (MA) resolved any disagreement and ensured consensus. Data regarding the type of studies, aims, populations, comparisons, and results were extracted into a Microsoft Excel^® 365 (2016) spreadsheet.

Study risk of bias assessment

The included studies were assessed for risk of bias using the Newcastle-Ottawa Scale (NOS) tool, which is validated for cohort studies and has been adapted for cross-sectional studies [12, 13]. NOS evaluates key domains, including selection of study groups, comparability between groups, and the ascertainment of exposure and outcomes, thereby allowing the assessment of selection and measurement biases. The tool uses a star rating system, where longitudinal studies of the highest quality can earn a maximum of nine stars, while cross-sectional studies can achieve up to ten stars. Methodological quality is categorized as poor, moderate, or high. Cohort studies scoring 7 or more points, and cross-sectional studies scoring 8 or more, are classified as high quality [14]. A score less than five indicates a weak study.

Vascular access questionnaire

All included studies used the Vascular Access Questionnaire (VAQ)—either the original version or the shorter form (SF-VAQ)—to assess patients’ perspectives and satisfaction with vascular access (Tables 1 and 2 in supplement). The original VAQ has 17 questions specified for HD patients, and each item is assessed by a 5-point (0–4) Likert scale, where 4 indicates the highest dissatisfaction.

The SF-VAQ is a shorter version of the original VAQ and includes three domains: physical symptoms, social functioning, and dialysis-related complications. A seven-point (1–7) Likert scale is utilized to assess each item within these domains, with 1 indicating the highest level of satisfaction and 7 indicating the highest level of dissatisfaction. Notably, the SF-VAQ has an additional item requiring patients to assess their overall satisfaction with a score of 1 signifying the highest level of dissatisfaction.

Statistical analysis

For the meta-analysis, only studies that utilized the SF-VAQ were included, ensuring methodological consistency in scale and structure to assess differences in patient-reported outcomes across various types of VA. These outcomes, which included overall satisfaction, physical symptoms, social functioning, and dialysis-related complications, were compared across three vascular access groups: AVF, CVC, and AVG.

To enable quantitative synthesis and comparability across studies, we standardized all Likert-scale scores to a 0–100 scale. For example, for studies that reported outcomes using a 7-point Likert scale, each score was linearly transformed using the formula:

It is important to note that in the SF-VAQ scoring system, a higher score in the overall satisfaction domain indicates greater satisfaction (i.e., lower dissatisfaction). In contrast, higher scores in the physical symptoms, social functioning, and dialysis-related complications domains reflect greater dissatisfaction and lower satisfaction. This directional difference was carefully considered when interpreting and pooling results across studies.

Three pairwise comparisons were performed: AVF vs. CVC, AVF vs. AVG, and CVC vs. AVG. For each comparison, the pooled mean difference and corresponding 95% confidence intervals (CIs) were estimated using a random-effects model with the Restricted Maximum Likelihood (REML) method, which provides an unbiased estimate of between-study variance (τ²). Heterogeneity among studies was assessed using Cochran’s Q test, the I² statistic, and H². I² values of 25%, 50%, and 75% were interpreted as indicative of low, moderate, and high heterogeneity, respectively.

To assess the robustness of our findings, sensitivity analyses were conducted by including and excluding specific studies based on the measurement scale. To assess publication bias, funnel plot symmetry was visually inspected, and Egger’s regression test was performed to detect small-study effects. If publication bias was identified, the Trim-and-Fill method was applied to estimate the number of potentially missing studies and to adjust the pooled effect size accordingly.

A Galbraith plot was used to further assess heterogeneity and to identify potential outlier studies. All statistical analyses were conducted using Stata version 18 (StataCorp, College Station, TX, USA). A two-tailed p-value < 0.05 was considered statistically significant.

Results

After excluding duplicates, 1253 studies were identified. Of these records, 11 studies fulfilled the inclusion criteria, giving a total of 2910 patients. The flow diagram is presented in Fig. 1. Ten of these studies were cross-sectional, while one was a prospective cohort (Table 1). All studies were conducted in high-income or upper-middle-income countries, and five of them were published in 2024. The number of patients in these studies ranged from 77 to 749, while the mean age varied from 53.5 to 70.3 years.

Fig. 1.

The flow diagram which included database searches, title and abstract screening, and eligibility assessment of records

Table 1.

Summary of final articles included in this study

Author, year of publication	Country	Study design	Population	Stated aims	Type of Questionnaire (No. of items), Measurement	Comparison	Associated conditions	Results
Wu et al. [15] 2024	China	Prospective cohort	229 pts: Pts with AVF (n = 194), AVG (n = 4), CVC (n = 31)	To explore the impact of VA satisfaction on HRQoL	SF-VAQ (13) 7-point Likert	AVF vs. AVG vs. CVC	• Overall mean duration of HD at baseline: 6 years	• Non-significant difference in overall satisfaction (measured by satisfaction item) of VA among pts with AVF, AVG, and TCC, both at baseline and after 2-year follow-up. • Pts with TCC had significantly worse satisfaction in the social and complications domains, compared to AVF, and AVG. • After a 2-year follow-up, a significant decrease in all domains of VAQ (physical, social, and complications) was found in AVF and TCC pts (not in pts with AVG). • Pts satisfied with VA (total VAQ score in all items < 7), retained HRQoL after a 2-year follow-up, but those dissatisfied with VA (total VAQ score ≥ 7) had a significant decrease in all dimensions of HRQoL from baseline to the follow-up. • The VAQ overall score, social domain score, and complication domain score significantly influenced HRQoL.
Li et al. [16] 2024	China	Cross-sectional	252 pts: Pts with AVF (n = 212), AVG (n = 7), TCC (n = 33)	To evaluate the association between VA satisfaction, HRQOL, and depression	SF-VAQ (13) 7-point Likert	AVF vs. AVG vs. TCC	• Pts with AVF were significantly younger than pts with AVG and TCC.	• Non-significant difference in VA satisfaction item among AVF, AVG, and TCC pts. • TCC pts had significantly the highest satisfaction regarding pain and bleeding, compared to AVF and AVG pts. • Significantly higher dissatisfaction was found in TCC pts regarding bathing and VA function, compared to AVF and AVG pts. • Pts with AVG were significantly more worried about hospitalization, compared to AVF and TCC pts. • Higher scores of VAQ in overall satisfaction item and complications domain had a significant negative correlation with HRQoL. • In multivariable analysis, overall satisfaction item and the physical domain of VAQ were factors significantly influencing depression.
Khawaja et al. [17] 2024	United Kingdom	Cross-sectional	539 pts: 195 aneurysmal AVF and 344 non-aneurysmal AVF	To compare pts characteristics between aneurysmal and non-aneurysmal AVFs	VAQ (17) 5-point Likert	Aneurysmal AVF vs. non-aneurysmal AVF	• Median duration of HD: aneurysmal (7 years), non-aneurysmal (3 years) •Median Duration of current access: aneurysmal (6 years) vs. non-aneurysmal (2 years) • Previous CVC: 58%	• Diabetic pts significantly had lower rates of aneurysmal AVF. • Pts with aneurysmal AVF had a significantly longer duration of HD and current access. • Total VAQ score did not differ significantly between aneurysmal and non-aneurysmal AVFs (5.22 vs. 4.99) or between different types of aneurysmal AVFs. • Aneurysmal AVFs had significantly better (lower) scores for bruising and clotting components of VAQ. • Aneurysmal AVFs had significantly worse (higher) scores for the appearance component of the VAQ. • Aneurysmal AVFs had significantly higher levels of satisfaction and ease of use.
Field et al. [18] 2024	United Kingdom	Cross-sectional	344 pts: 13 pts with AVF created endovascularly (EndoAVF) and 331 pts with AVF created surgically	To compare the perspective of pts about EndoAVF and surgically created AVF	VAQ (17) 5-point Likert	EndoAVF vs. surgical AVF	• Median Duration of HD: EndoAVF (3 years) vs. surgical AVF (2 years) • Median Duration of current access: EndoAVF (2 years), surgical AVF (2 years)	• Total VAQ score did not differ significantly between EndoAVF and surgical AVF (4.5 vs. 5.2) • EndoAVF had a significantly elevated pain score compared to the surgical AVF (mean 0.46 vs. 0.28, p = 0.040) • high levels of satisfaction and ease of use were reported by EndoAVF and surgical AVF, without significant difference.
Barros et al. [19] 2024	Portugal	Cross-sectional	156 pts: Pts with AVF (n = 129), AVG (n = 9), CVC (n = 18)	To evaluate satisfaction levels related to VA type	SF-VAQ (13) 7-point Likert	AVF vs. AVG vs. CVC	• Previous CVC: 53%	• High satisfaction with all VA types (measured by overall satisfaction item): mean AVF, AVG, and CVC scores were 5.44, 4.89, and 4.72 (p = 0.163). • Significantly AVG was associated with higher levels of pain (4 vs. 2.94 for AVF and 1.22 for CVC, p < 0.001) • Significantly better physical appearance was obtained for AVF (2.7 vs. 3.89 for AVG and 4.33 for CVC, p = 0.017) • CVC significantly interfered more with bathing than AVF and AVG. • Concerns about hospitalization due to VA are significantly greater for AVG and CVC than for AVF. • Pts with VA in the dominant arm are more concerned with the longevity of VA. • Pts with previous CVC had significantly higher scores for pain. • Age below 60, female gender, and pts without previous VA failure are independent predictors of worse VAQ score.
Maldonado et al. [20] 2023	Spain	Cross-sectional	91 pts: 55 AVF and 36 CVC pts	To evaluate the quality of life of pts based on the VA type	SF-VAQ (13) 7-point Likert	AVF vs. CVC	• Mean duration of HD: AVF (5.8 years) vs. CVC (3 years) (p < 0.05) • Median Duration of current access: AVF (4.9 years) vs. CVC (1.8 years) (p < 0.05) • Start HD with CVC: 76%	• Mean VAQ scores did not differ significantly (2.4 for AVF vs. 2.1 for CVC) • Social and complication domains of VAQ had significantly better results for AVF. The physical domain was better for CVC but did not significantly differ (0.053). • AVF pts were significantly more satisfied with hygiene (p = 0.0001). • AVF pts had RR = 9.7 (p = 0.012) more dissatisfied with bleeding of VA. • Dominant arm AVF was significantly more satisfactory in the social domain compared to non-dominant (p = 0.0001) • Males who had both AVF and CVC significantly preferred AVF compared to females. • Females were more concerned than males about hospitalization and longevity of VA. • A weak correlation between the social domain of VAQ and age was found (r: -0.255, p = 0.017)
Maguire et al. [21] 2022	Ireland	Cross-sectional	119 pts: 62 AVF and 57 CVC pts	To assess patient satisfaction relating to vascular access	SF-VAQ (13) 7-point Likert	AVF vs. CVC		• In overall satisfaction item, AVF pts had significantly greater satisfaction, compared to CVC pts(6.2 vs. 5, p = 0.008). • Overall satisfaction with the physical domain was high, although AVF pts expressed significantly more problems with bleeding, swelling, bruising, and in the overall physical domain. • AVF pts reported significantly greater interference of VA with daily activities and appearance whereas CVC pts expressed more difficulties in bathing. • Among the complications domain, the only difference was found to be greater difficulties in caring for VA among CVC pts. • The sum of scores in physical, social, and complication domains were higher in AVF pts but did not differ significantly from CVC pts (p = 0.07). • After adjusting age, sex, and comorbidities, CVC resulted in significantly lower difficulties in the physical domain (but not significantly in the social and complications domain)
Field et al. [22] 2019	United Kingdom	Cross-sectional	749 pts: Pts with AVF (n = 539), AVG (n = 34), CVC (n = 174), both CVC and AVF (n = 2)	to identify pts characteristics influencing the perception and outcomes of VA	VAQ (17) 5-point Likert	AVF vs. AVG vs. CVC	• Overall median duration of HD: 3 years • Previous CVC: 67.4%	• In total VAQ score (sums of all items), AVF pts resulted in significantly greater satisfaction, compared to CVC and AVG. • Pts with CVC had significantly better scores regarding bleeding, bruising, and swelling. • Pts with AVF had significantly better results regarding infection, clotting, VA function, protecting VA, interfering with daily activities, and worries about hospitalization and longevity. • Multivariable analysis revealed that only younger age, female gender, having cardiac disease, and radiology intervention in the prior year are predictors of high VAQ score. Type of VA was not a significant variable. • VAQ score did not differ between dominant arm AVF/AVG and non-dominant. • Pts with brachiobasilic AVF had significantly more problems with sleeping and worries about its function, compared to brachiocephalic and radiocephalic.
Sridharan et al. [23] 2017	United States	Cross-sectional	77 pts: Pts with AVF (n = 48), AVG (n = 11), TDC (n = 18)	To evaluate the impact of HD access type on patient satisfaction	SF-VAQ (13) 7-point Likert	AVF vs. AVG vs. TDC	First VA: TDC in 61% of pts	• In the overall satisfaction item, AVF resulted in significantly greater satisfaction (4.5 vs. 6.5 for TDC vs. 7 for AVG). • Multivariate analysis showed that less than one year on HD, number of access-related hospitalizations, AVG, and TDC are independent predictors of high VAQ score. (Age, sex, and previous TDC were not significant variables.) • Pts with TDC were significantly more satisfied in the physical domain. • TDC resulted in more dissatisfaction with the complications domain. • Both AVG and TDC pts were significantly more dissatisfied with the social domain. • Bleeding bothered pts with AVG significantly more than other VAs. •Pts with AVG significantly required more operative revisions.
Kosa et al. [24] 2015	Canada	Cross-sectional	132 pts: Pts with AVF (n = 35), AVG (n = 14), CVC (n = 83)	To develop and evaluate psychometrics of SF-VAQ	SF-VAQ (13) 7-point Likert	AVF vs. AVG vs. CVC	Not applicable	• The satisfaction of pts (measured by overall satisfaction item) with AVF was greater than CVC and AVG. • In the physical domain, pts with AVG were more dissatisfied, compared to CVC and AVF. • In the social domain, pts with CVC were more dissatisfied, compared to AVF and AVG. • In the complications domain, pts with CVC were more dissatisfied, compared to AVF and AVG. • The overall VAQ score (sums of scores in all domains) showed greater satisfaction with AVF, followed by CVC and AVG.
Quinn et al. [25] 2008	Canada	Cross-sectional	222 pts: Pts with AVF (n = 118), CVC (n = 104)	To develop VAQ and measure the perspectives of pts regarding VA	VAQ (17) 5-point Likert	AVF vs. CVC	• Median duration of HD: AVF (3.2 years) vs. CVC (1.4 years) (p < 0.01)	• Mean VAQ scores did not differ significantly between AVF and CVC pts (6.2 for AVF vs. 5.4 for CVC, p = 0.36). • Access type was not a significant predictor of VAQ scores in either the unadjusted or adjusted models. • Pts with AVF were significantly more bothered by pain, bleeding, bruising, swelling, and appearance of VA, compared to CVC pts. • Age was found to modify the relationship between VA type and VAQ scores (interaction term p < 0.01); pts younger than 65 had lower VAQ scores with AVFs, while older pts had lower VAQ scores with CVCs. • Non-significant trend toward higher rates of VA related procedure in CVC pts (62% vs. 48% in AVF, p = 0.06) • Non-significant trend toward increased hospitalization in CVC pts (7% vs. 2% in AVF, p = 0.06)

AVF, Arteriovenous fistula; AVG, arteriovenous graft; CVC, central venous catheter; TDC, tunneled dialysis catheter; TCC, tunneled cuffed catheter; Pts, patients; HRQoL, health-related quality of life; VAQ, vascular access questionnaire; SF-VAQ, short-form VAQ; VA, vascular access; HD, hemodialysis

Nine studies assessed patient satisfaction with various types of VA [15, 16, 19–25]. Six of these studies [15, 16, 19, 22–24] compared AVF (total number of patients (n) = 1157), AVG (n = 79), and CVC (n = 357), while the remaining studies [20, 21, 25] focused solely on comparing AVF (n = 235) with CVC (n = 197). Seven studies used the short-form VAQ (SF-VAQ), whereas the remaining articles used the original VAQ.

Overall satisfaction with different types of VA

A meta-analysis was conducted to compare overall satisfaction scores, as measured by the SF-VAQ, between patients with AVF and those with CVC, based on seven studies (n = 1011) [15, 16, 19–21, 23, 24]. Given the observed heterogeneity (I² = 60.30%, τ² = 17.36, Q(6) = 13.99, p for heterogeneity = 0.03) and observations from the Galbraith Plot, a random-effects model was applied to account for between-study variability. The Restricted Maximum Likelihood (REML) method was selected as the most statistically robust estimator for τ², ensuring an unbiased variance estimate. The pooled mean difference (MD) from seven studies was 7.06% (95% CI: 2.98 to 11.14), indicating a significantly higher overall satisfaction score in AVF patients compared to CVC patients (p = 0.0007) (Fig. 2). Furthermore, funnel plot inspection revealed asymmetry, and the Egger test confirmed the presence of small-study effects (β = 4.72, SE = 1.46, z = 3.22, p = 0.0013).

Fig. 2.

Meta-analysis of overall satisfaction difference among AVF and CVC patients, as measured by SF-VAQ. Higher mean scores indicate more satisfaction. The size of the boxes indicates weight of the study in proportion to the pooled estimate. The size of the diamond represents the overall pooled effects, and the width of the diamond represents the 95% CI of the point estimate of the pooled effect. AVF; arteriovenous fistula, CVC; central vein catheter, REML; Restricted Maximum Likelihood

In the meta-analysis comparing AVF and AVG, the pooled overall satisfaction score from five studies (n = 663) was higher in AVF patients; however, the difference was not statistically significant (MD: 5.12%, 95% CI: -0.42 to 10.66, p = 0.069) (Fig. 3). Between CVC and AVG (five studies, n = 228), no significant difference in the pooled overall satisfaction score was noted (p = 0.99) (supplement).

Fig. 3.

Meta-analysis of overall satisfaction difference among AVF and AVG patients, as measured by SF-VAQ. The size of the boxes indicates weight of the study in proportion to the pooled estimate. Higher mean scores indicate more satisfaction. The size of the diamond represents the overall pooled effects, and the width of the diamond represents the 95% CI of the point estimate of the pooled effect. AVF; arteriovenous fistula, AVG; arteriovenous graft, REML; Restricted Maximum Likelihood

In the sensitivity analysis, we included one additional study that used the original VAQ [22] (after transforming scores to a 0–100 scale) (supplement). When this study was added to the meta-analysis, the difference in overall satisfaction between AVF and AVG became statistically significant (MD: 3.49, 95% CI: 1.38–5.60, p = 0.0012; I² = 0.28%, p for heterogeneity = 0.43), whereas it was not significant in the main analysis. The significance of the differences in satisfaction between AVF and CVC, and between CVC and AVG, remained unchanged in the sensitivity analysis.

VA types and physical symptoms

The pooled overall dissatisfaction in the domain of physical symptoms (seven studies, n = 1011) was significantly higher among patients with AVF, compared to CVC (MD: 8.76%, 95% CI: 4.19 to 13.34, p = 0.0002) (Fig. 4). Additionally, patients with CVC also showed lower overall dissatisfaction in this domain, compared to AVG patients (MD: -13.78%, 95% CI: -21.61 to -5.95, p = 0.006) (supplement). In this domain, no significant difference was observed between AVF and AVG patients (supplement).

Fig. 4.

Meta-analysis of physical symptoms overall dissatisfaction difference among AVF and CVC patients, as measured by SF-VAQ. Higher mean scores indicate more dissatisfaction. The size of the boxes indicates weight of the study in proportion to the pooled estimate. The size of the diamond represents the overall pooled effects, and the width of the diamond represents the 95% CI of the point estimate of the pooled effect. AVF; arteriovenous fistula, CVC; central vein catheter, REML; Restricted Maximum Likelihood

VA types and social functioning

The meta-analysis of seven studies (n = 1011) revealed that patients with AVF had lower overall dissatisfaction in the social functioning domain, compared to CVC patients (MD: -9.48%, 95% CI: -15.98 to -2.98; p = 0.004) (Fig. 5). In this domain, no significant differences were observed between AVF and AVG patients, or between AVG and CVC patients (supplement).

Fig. 5.

Meta-analysis of social functioning overall dissatisfaction difference among AVF and CVC patients, as measured by SF-VAQ. Higher mean scores indicate more dissatisfaction. The size of the boxes indicates weight of the study in proportion to the pooled estimate. The size of the diamond represents the overall pooled effects, and the width of the diamond represents the 95% CI of the point estimate of the pooled effect. AVF; arteriovenous fistula, CVC; central vein catheter, REML; Restricted Maximum Likelihood

Most studies using the SF-VAQ reported that CVCs interfere significantly more than AVFs and AVGs with bathing or showering [16, 19, 21, 23, 24]. Among studies specifically assessing the physical appearance of each type of VA, two studies [21, 25] found that AVFs caused more dissatisfaction with physical appearance than CVCs. However, Barros et al. reported significantly greater satisfaction with the physical appearance for AVFs [19]. Additionally, Li et al. and Field et al. found no significant differences in physical appearance satisfaction among VA types [16, 22].

Two studies reported that AVFs interfered more with daily activities compared to other VA types [16, 21]; however, Field et al. found greater satisfaction with AVFs [22]. In the study by Sridharan et al., patients with AVG had the lowest satisfaction in daily activities among all VA types [23].

VA types and dialysis-related complications

The meta-analysis of seven studies (n = 1011) revealed that, compared to patients with CVC, those with AVF reported significantly lower dissatisfaction related to dialysis complications (MD: -10.08%, 95% CI: -12.66 to -7.50; p < 0.0001) (Fig. 6). Additionally, patients with AVF reported lower dissatisfaction in this domain compared to those with AVG (MD: -5.01%, 95% CI: -9.98 to -0.04; p = 0.05), but there was no significant difference between AVG and CVC patients (Supplement).

Fig. 6.

Meta-analysis of dialysis-related complications overall dissatisfaction difference among AVF and CVC patients, as measured by SF-VAQ. Higher mean scores indicate more dissatisfaction. The size of the boxes indicates weight of the study in proportion to the pooled estimate. The size of the diamond represents the overall pooled effects, and the width of the diamond represents the 95% CI of the point estimate of the pooled effect. AVF; arteriovenous fistula, CVC; central vein catheter, REML; Restricted Maximum Likelihood

AVF creation and aneurysms

One study evaluated patient satisfaction with AVF created by endovascular technique (EndoAVF) compared to surgically created AVF [18]. Although pain perception was significantly higher in the EndoAVF group, overall satisfaction was high in both groups, and the VAQ scores did not differ significantly.

Another study compared aneurysmal and non-aneurysmal AVF from the patient’s perspective [17]. Patients with aneurysmal AVF, albeit more worried about VA appearance, had better scores (higher satisfaction) regarding bruising and clotting associated with AVF. The overall VAQ scores did not show a significant difference between the two groups; however, patients with aneurysmal AVFs reported significantly higher levels of satisfaction and ease of use when asked separately from the VAQ questionnaire.

Age and VA satisfaction

Various associations were found for age and VAQ score. Two studies found that younger age is a significant independent predictor of worse VAQ scores [19, 22]. The study by Maldonado et al. demonstrated a weak negative correlation between age and the score of only the social domain of SF-VAQ (not physical or dialysis-related complications domains) [20]. One study showed that patients younger than 65 have better VAQ scores with AVFs, but older patients have more satisfying VAQ scores with CVCs [25]. Sridharan et al. found no significant association between age and VAQ score after adjusting other variables such as VA type [23].

Gender and VA satisfaction

Two studies showed that in multivariable analysis, female gender is significantly associated with worse VAQ scores [19, 22]. In one study, females were found to be more worried about hospitalization and the longevity of VA than men [20]. Additionally, among patients with a previous history of both CVC and AVF, men tended to opt for AVF significantly more than women [20]. In contrast, the study by Sridharan et al. did not identify gender as a significant predictor of high VAQ scores [23].

Location of AVF and VA satisfaction

Field et al. reported no significant differences between AVFs or AVGs placed on the dominant arm compared to the non-dominant arm, either in the mean VAQ score or in any individual VAQ items [22]. Although Barros et al. found that patients with AVFs on the dominant arm are primarily more concerned about the longevity of the VA [19], interestingly, Maldonado et al. revealed that AVFs on the dominant arm are significantly more satisfactory in the social functioning domain [20]. Regarding the location of the AVF in the arm, brachiobasilic AVFs were significantly more associated with sleep disturbances and concerns about their function compared to brachiocephalic and radiocephalic AVFs [22].

Risk of bias assessment

The NOS bias assessment showed that the single cohort study received a score of 7 out of 9, indicating high methodological quality (Table 3 in supplement). The remaining cross-sectional studies were evaluated on a 10-point scale, with scores ranging from 4 to 8. Of these, eight studies were rated as high quality (scores ≥ 8), and three were rated as moderate quality (scores of 7). One study received a score of 4, indicating poor methodological quality.

Discussion

To the best of our knowledge, this is the first systematic review and meta-analysis to include studies assessing the satisfaction of HD patients with different VAs. The most important finding of this review is the overall higher satisfaction associated with AVFs compared to other vascular accesses. Although AVF patients reported more discomfort related to physical symptoms, this was offset by significantly greater satisfaction in the domains of social functioning and dialysis-related complications, contributing to higher overall satisfaction.

VA decisions often reflect clinicians’ views of the “optimal access” or national policy goals, with limited attention to patients’ perspectives. However, evidence increasingly shows that for many with chronic illness, quality of life outweighs concerns about complications or even mortality [21, 26]. Patient-reported outcomes and satisfaction are now recognized as critical, with studies like Elamin et al. showing that patients’ perceived health correlates with survival [27]. Satisfaction with VA—an important aspect of HD quality of life—has also been linked to survival outcomes [8]. Understanding patients’ perspectives on their health, quality of life, and, in the case of HD patients, satisfaction with VA, is increasingly recognized as essential for clinical monitoring, enhancing healthcare services, promoting adherence to medical recommendations, and potentially influencing survival outcomes [24–26, 28].

Through checking the eligibility criteria of identified records, we noticed that the best way to capture the perspectives of HD patients on satisfaction with VA is by utilizing a questionnaire specially designed and validated for this purpose. The VAQ offers a structured mechanism for identifying and scoring patient perceptions in this context, with an estimated completion time of 4 to 10 min [15, 16, 23, 29]. The VAQ was originally developed and validated in Canadian HD patients and found to have high test-retest reliability and internal consistency in psychometric evaluation [25].

Generic quality-of-life tools like the SF-36, HUI3, and Euro-QOL are widely used in HD but lack specificity for assessing VA care [24, 30]. Even disease-specific tools like the KDQOL and KDHRQOL include few VA-related items and may be burdensome due to their length [31]. Focusing on patient satisfaction with VA, rather than the overall quality of life, may be more effective for improving the delivery of sustainable VA programs and quality improvement initiatives [21].

AVF has long been regarded clinically as the VA of choice for HD due to its superior patency, lower risk of infectious complications, and association with reduced all-cause mortality compared to AVG or CVC [23, 32]. Nevertheless, from the patients’ perspective, the creation of an AVF is often seen as an inflection point in their lives, primarily due to the associated feelings of abnormality in the body [10]. Furthermore, our review revealed that patients with AVF generally experience more physical symptoms such as pain and bleeding [16, 19, 21–25]. Although patients with AVF may not necessarily view their body appearance as impaired, the majority report being bothered by their VA in daily life activities [16, 19, 21, 22]. Despite this, AVFs are associated with significantly fewer dialysis-related complications, such as concerns about care, hospitalization, longevity, or function. These perspectives from previous patients are worth addressing at the start of HD consultations for new beginners, with a strong emphasis on individual needs, conditions, and the importance of each aspect.

The start of dialysis can be particularly frustrating for new patients, whereas experienced patients may become desensitized over time. This is reflected in the study by Sridharan et al., where < 1 year on dialysis was identified as a predictor of dissatisfaction [23]. However, as indicated by VAQ scores, several studies reported that patient satisfaction with their vascular access increases with both patient age and the age of the access [19, 20, 22, 23, 33]. Remarkably, even in a two-year follow-up, all VAQ domains showed significant improvement in patients with AVFs or CVCs [15]. This suggests that patients may go through an adjustment period as they become accustomed to their vascular access, which could be an important consideration for counseling before they begin their access journey.

Our review revealed that 53–76% of HD patients reported previous use of CVC [17, 19, 20, 22, 23]. Furthermore, the use of CVCs is significantly higher among new patients, compared to those more seasoned [21]. In a study of 599 chronic HD patients, Graham et al. identified shorter dialysis duration as an independent predictor of CVC use [34]. They concluded that the high prevalence of CVC use is primarily attributed to patient-specific factors, such as comorbidities or unsuitable vessels, rather than resource limitations or the attitudes of nephrologists. Although the “fistula first” policy encourages and, when necessary, insists on AVF placement, nearly 75% of patients currently using CVCs were unwilling to switch to a new AVF [20]. While unscheduled initiation of HD may play a role in CVC use, no differences were found in the information provided to patients who started HD with a CVC versus an AVF [20].

Studies show that younger age and female gender are linked to greater dissatisfaction with VA [19, 22]. Younger patients, who are often dissatisfied with the social functioning domain of VA, may find greater satisfaction with AVFs compared to CVCs, which are reported to cause the least interference with social functioning [15, 19, 20, 25]. Female patients are more concerned with hospitalization and VA longevity, both of which tend to favor AVFs. Tailoring pre-dialysis counseling to address these concerns may improve satisfaction in these groups.

Elderly patients represent a distinct population to consider, with nearly half of HD patients being over the age of 65 [25]. With the added burden of multiple comorbidities and poor fistula maturation rates, coupled with the observation that overall satisfaction with VA tends to increase with age, these patients may benefit most from an access type that maximizes functionality while minimizing hospitalizations. As Quinn et al. portrayed, patients over 65 have better VAQ scores with CVCs; in some cases, a well-functioning CVC is more patient-centered than a poorly functioning or repeatedly revised AVF, which also carries higher risks of bleeding and bruising [23, 25, 35, 36].

Although aneurysmal AVF may deter some patients due to appearance, Khawaja et al. reported that it does not significantly impact long-term satisfaction, as patients value the ease of cannulation [17]. Besides, endovascular AVF offers an alternative approach for patients who are apprehensive about or unwilling to undergo further surgical access creation. It provides comparable satisfaction to surgical AVF from the patients’ perspective [18]. Highlighting these benefits during consent and outpatient discussions could improve patient acceptance of AVF and promote the “fistula first” policy.

Maldonado et al. found significantly higher satisfaction with AVFs placed in the dominant arm (p = 0.0001), a trend also observed by Barros et al., though not statistically significant (p = 0.142) [19, 20]. This may reflect improved outcomes after prior failures in the other arm. As the dominant arm often offers better vessel size for maturation [22], the routine preference for the non-dominant arm should be reconsidered based on ultrasound findings and individual patient needs.

Limitations

This review has several limitations. First, the protocol was not registered in the PROSPERO database prior to commencement. Second, although all included studies used the VAQ questionnaire, structural and scaling differences exist between the original VAQ and the SF-VAQ. To maintain methodological consistency, only studies using the SF-VAQ were included in the primary meta-analysis. However, a sensitivity analysis that incorporated one study using the original VAQ (after score transformation) showed a statistically significant difference in satisfaction between AVF and AVG, suggesting that findings related to this comparison may be influenced by measurement differences. This highlights the need for standardized patient-reported outcome tools in future research.

Additionally, the predominance of cross-sectional studies limits causal inferences. Most studies were conducted in high- or upper-middle-income countries, leaving the perspectives of patients in low-income settings underrepresented. Use of self-reported data introduces potential recall bias, and variable response rates (79.5–93%) may affect generalizability.

Conclusion

This review shows that AVFs are generally associated with higher overall patient satisfaction than other vascular access types, despite some physical discomfort. While AVF patients may experience more physical discomfort, this is outweighed by significantly greater satisfaction in areas such as social functioning and dialysis-related complications. These findings underscore the value of incorporating patient-reported outcomes into vascular access decisions. Integrating clinical factors—such as anatomy, age, gender, and comorbidities—with patient concerns can guide surgeons in offering access options that best support quality of life.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HD

Hemodialysis

DALY

Disability-adjusted life-years

KRT

Kidney replacement therapy

VA

Vascular access

AVF

Arteriovenous fistula

AVG

Arteriovenous graft

CVC

Central venous catheter

NOS

Newcastle-Ottawa Scale

VAQ

Vascular access questionnaire

SF-VAQ

Short-form VAQ

CI

Confidence interval

REML

Restricted Maximum Likelihood

MD

Mean difference

Author contributions

A.A. and S.P. wrote the main manuscript text and contributed to conceptualization, methodology, validation and data curation. J.S. contributed to methodology, supervision, validation, and manuscript review. M.A. supported methodology, supervision, validation, and manuscript editing. M.Y. performed formal analysis and contributed to conceptualization, methodology, software, and supervision. All authors reviewed and approved the final manuscript. All authors are designated as corresponding authors.

Funding

This research received no funding.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request. Additional data is provided within the supplementary.

Declarations

Ethical approval

Ethical approval was not required as this study involved analysis of published data.

Consent for publication

Not applicable.

Registration and protocol

Not registered.

Competing interests

The authors declare no competing interests.