Introduction: Vascular access usage varies widely across countries. Previous studies have evaluated the association of clinical outcomes with the three types of vascular access, namely, arteriovenous fistula (AVF), arteriovenous graft (AVG), and tunneled and cuffed central venous catheter (TC-CVC). However, little is known regarding the association between arterial superficialization (AS) and the mortality of patients. Methods: A nationwide cohort study was conducted using data from the Japanese Society for Dialysis Therapy Renal Data Registry (2006–2007). We included patients aged ≥20 years undergoing hemodialysis with a dialysis vintage ≥6 months. The exposures of interest were the four types of vascular access: AVF, AVG, AS, and TC-CVC. Cox proportional hazard models were used to evaluate the associations of vascular access types with 1-year all-cause and cause-specific mortality. Results: A total of 183,490 maintenance hemodialysis patients were included: 90.7% with AVF, 6.9% with AVG, 2.0% with AS, and 0.4% with TC-CVC. During the 1-year follow-up period, 13,798 patients died. Compared to patients with AVF, those with AVG, AS, and TC-CVC had a significantly higher risk of all-cause mortality after adjustment for confounding factors: adjusted hazard ratios (95% confidence intervals) – 1.30 (1.20–1.41), 1.56 (1.39–1.76), and 2.15 (1.77–2.61), respectively. Similar results were obtained for infection-related and cardiovascular mortality. Conclusion: This nationwide cohort study conducted in Japan suggested that AVF usage may have the lowest risk of all-cause mortality. The study also suggested that the usage of AS may be associated with better survival rates compared to those of TC-CVC in patients who are not suitable for AVF or AVG.

End-stage kidney disease (ESKD) remains a major public health concern worldwide [1]. Globally, the number of patients receiving kidney replacement therapy exceeded 2.6 million and is projected to double to 5.4 million by 2030 [2]. The majority of these patients are treated with hemodialysis [2]. Despite significant therapeutic advances, patients undergoing hemodialysis have a higher risk of complications, such as cardiovascular disease and infection [3]. Among hemodialysis practices, vascular access is one of the major causes of morbidity and mortality [4, 5]. Moreover, the cost of access-related complications is extremely high [6]. Therefore, the creation and maintenance of reliable vascular access are crucial in this population [7, 8].

Arteriovenous fistula (AVF) is the most common and preferred vascular access option due to fewer associated complications [9, 10]. The Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines for vascular access discourage the use of a central venous catheter (CVC) for long-term hemodialysis access [10]. Prolonged CVC use is associated with increased risks of infection, hospitalization, and mortality [11, 15] and requires frequent replacement due to infection or dysfunction [16]. Moreover, the KDOQI guidelines recommend AVF over arteriovenous graft (AVG) [10]. A meta-analysis of cohort studies showed that AVF usage was associated with a lower risk of serious infections and mortality than AVG usage [17]. However, several studies have shown conflicting results regarding the optimal vascular access types based on mortality, especially in elderly patients [18, 21]. Therefore, further studies are required to determine whether AVF should be the preferred vascular access for patients requiring hemodialysis.

There is wide variation in practice patterns regarding vascular access worldwide [22]. Although the majority of patients undergo hemodialysis using AVF, data from the Dialysis Outcomes and Practice Patterns Study (DOPPS) revealed that AVF usage in patients undergoing maintenance hemodialysis ranged from 49% to 92%, whereas tunneled and cuffed CVC (TC-CVC) usage ranged from 1% to 45% across countries [22]. There are also differences in AVF maturation time and AVF patency internationally [23, 24]. Overall, Japanese patients undergoing hemodialysis with low mortality rates have a higher prevalence of AVF usage, shorter time to first AVF cannulation, and longer AVF survival compared with those in other DOPPS countries [22, 24]. In addition to the three aforementioned types of vascular access, arterial superficialization (AS), a vascular access that mainly superficializes the brachial artery and uses it as an outflow route, is widely used in Japan (more information is presented in Methods) [9]. The Japanese Society for Dialysis Therapy (JSDT) guidelines recommend AS or TC-CVC as alternative vascular access, especially among patients for whom AVF or AVG is not suitable (online suppl. Fig. 1; for all online suppl. material, see www.karger.com/doi/10.1159/000529991) [9]. However, little is known regarding whether AS usage is associated with a reduced risk of mortality compared with the usage of TC-CVC [25, 26]. Moreover, considering that AS neither includes artificial materials nor affects cardiac output [27], the survival of patients with AS may not be inferior to that of patients with AVG. Therefore, comprehensive studies comparing mortality with the four types of vascular access are needed.

In the present study, we examined the association of mortality with the four types of vascular access (AVF, AVG, AS, and TC-CVC) using a national database in Japan, in which patients undergoing hemodialysis have high survival rates. We hypothesized that AS usage would be associated with a decreased risk of mortality, at least compared with TC-CVC usage.

Database

All data used in this study were retrieved from the JSDT Renal Data Registry (JRDR) database. The detailed methods associated with JRDR have been described previously [28]. JSDT started to conduct annual questionnaire surveys of dialysis facilities across Japan in 1968. The questionnaires are filled out by medical staff in each facility and are sent back to the JSDT to create a database and examine national trends in dialysis care. Given that the response rate to the questionnaire was 98.4% in 2006 (3,985 of 4,051 facilities for chronic dialysis) [28], the database included nearly all dialysis facilities in Japan.

Study Design

We conducted a 1-year retrospective cohort study using the JRDR data collected as of December 31, 2006, and December 31, 2007. The follow-up period started on December 31, 2006, and ended on events involving death, withdrawal of dialysis, kidney transplantation, or on December 31, 2007, whichever was earliest. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting our results [29].

Study Participants

We included patients (i) aged ≥20 years, (ii) who received maintenance hemodialysis, and (iii) with a dialysis vintage ≥6 months [30]. The exclusion criteria were as follows: (i) home hemodialysis; (ii) acute phase of myocardial infarction, cerebral infarction, or cerebral hemorrhage; (iii) unknown or missing information regarding vascular access; and (iv) loss to follow-up.

Type of Vascular Access

The exposure of interest was the type of vascular access. Patients were categorized into four groups based on the vascular access type: AVF, AVG, AS, and TC-CVC. Following the usage of vascular access at the time of study enrollment, they remained exposed regardless of whether the vascular access changed or not. We obtained data on the exposure from the JRDR database in 2006 [28].

AS has been widely used for decades as one of the most common types of vascular access in Japan [9]. Unlike AVF and AVG, AS does not affect the cardiac output because it only superficializes the brachial or femoral arteries to subcutaneous tissue without vascular anastomosis [27]. Thus, the main indications for AS are heart failure (ejection fraction ≤40%), central venous stenosis, or peripheral circulatory disorders (online suppl. Fig. 1). However, as the AS is used as an outflow route, cannulation of both the superficialized artery and the superficial vein for blood return is required for each hemodialysis session. The JSDT guidelines recommend the brachial artery as the first choice because of its ease of operation and few complications, and suggest the AS and TC-CVC as alternative options of vascular access, especially for patients who are not suitable for AVF and AVG creation [9].

Covariates

The following information was retrieved from the JRDR database in 2006: age, sex, body mass index (BMI), dialysis vintage (in years), duration of hemodialysis treatment (hours per week), cause of ESKD, history of myocardial infarction, cerebral infarction, cerebral hemorrhage, limb amputation, and laboratory measurements (pre-dialysis hemoglobin, albumin, calcium, phosphate, intact parathyroid hormone [PTH], and ferritin), and the within-facility proportion of patients with AVF [28]. If the serum albumin levels were <4.0 g/dL, the serum calcium levels were corrected as follows: corrected calcium (mg/dL) = measured calcium (mg/dL) + [4.0 – serum albumin (g/dL)].

Outcomes

The primary outcome of interest was all-cause mortality during the 1-year follow-up period. The secondary outcomes were infection-related mortality and cardiovascular mortality. Infection-related deaths were defined as deaths resulting from sepsis, central nerve infection, influenza, pneumonia, gastrointestinal infection, viral hepatitis, peritonitis, urinary tract infection, tuberculosis, human immunodeficiency virus infection, and other infections. Cardiovascular deaths included those due to cardiac diseases (heart failure, pulmonary edema, ischemic heart disease, arrhythmia, valvular heart disease, other heart diseases, or hyperkalemia/sudden death) or cerebrovascular diseases (subarachnoid hemorrhage, cerebral hemorrhage, cerebral infarction, or other cerebrovascular diseases). We collected data on the month of death as well as causes of death from the JRDR database in 2007 [31]. As the JRDR does not collect data regarding the day of death, withdrawal of dialysis, and kidney transplantation, we calculated the time of an event by assuming that each event occurred on the 15th day of the month [32]. Classification codes of the causes of death were reported previously [31]. The causes of death were diagnosed and reported by on-site physicians based on patients’ medical records.

Statistical Analysis

Patient characteristics based on the type of vascular access were expressed as means (standard deviation) or medians (interquartile range) for continuous variables and percentages for categorical variables. Differences in baseline characteristics between the four groups were tested using the χ2 test for categorical variables and one-way ANOVA or the Kruskal-Wallis rank sum tests for continuous variables. The Kaplan-Meier method was used to estimate survival, and survival differences according to the vascular access type were analyzed with the log-rank test. Cox proportional hazard model was used to estimate the association between the type of vascular access and mortality. We tested the proportional hazard assumptions of the Cox model using the log-log plot of survival and the Schoenfeld residuals. We adjusted for the following covariates: age, sex, BMI, cause of ESKD, dialysis vintage, hemodialysis duration, comorbidities (history of myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation of limbs), and laboratory data (hemoglobin, albumin, albumin-adjusted calcium [categorized as <8.4, 8.4–9.2, 9.3–10.0, 10.1–10.9, and ≥11.0 mg/dL], phosphate [categorized as <3.5, 3.5–4.7, 4.8–6.0, 6.1–7.3, and ≥7.4 mg/dL], intact PTH [categorized as <60, 60–180, 181–240, 241–499, and ≥500 mg/dL], and ferritin [categorized as <50, 50–99, 100–199, and ≥200 ng/mL]) [32, 33]. To account for different clinical practice patterns across facilities, we also adjusted for the proportion of AVF within the four types of vascular access (AVF, AVG, AS, and TC-CVC) in each facility. In addition to the cause-specific Cox hazard model described above, as other causes of death are considered competing risks in cause-specific mortality analyses, the Fine and Gray model was used to incorporate the rates of competing risks into the subdistribution hazard ratios (HRs) [34]. Patients were censored at the time of withdrawal of hemodialysis or kidney transplantation.

Subgroup analyses were conducted by age (<75 and ≥75 years), sex, dialysis vintage (<10 and ≥10 years), cause of ESKD (diabetes or others), relationship between vascular access at hemodialysis and study initiation (i.e., whether or not vascular access at the two time points for each patient was identical), history including myocardial infarction, cerebral infarction, cerebral hemorrhage, and limb amputation at baseline (presence or absence). Multiplicative interaction terms between the type of vascular access and each covariate were entered into the Cox proportional hazard model to evaluate the potential interaction on all-cause mortality.

We performed two sensitivity analyses to examine the robustness of our main results. First, we performed multiple imputation in the same multivariable Cox proportional hazard model and Fine and Gray model as above. The missing values of all covariates were imputed by chained equation method. Three covariates, which were skewed (dialysis vintage, intact PTH, and ferritin), were log-transformed to normalize the distribution before multiple imputation. We created five imputed datasets, which were analyzed separately and combined using Rubin rules. Subsequently, two propensity score matching analyses were performed to compare the risk of mortality between patients with AS and those with TC-CVC or AVG, respectively [35]. We estimated the propensity score by fitting a logistic regression model that adjusted for all covariates. We created 1:1 matches between the AS and TC-CVC or AVG groups, respectively, using nearest neighbor matching without replacement [35]. Propensity scores were matched within a caliper width of 20% of the standardized difference of the propensity scores. An absolute standardized difference of <10% was considered to denote negligible imbalances between the AS and TC-CVC or AVG groups, respectively [35]. After propensity score matching, HRs of all-cause mortality, as well as HRs and subdistribution HRs of cause-specific mortality, were compared.

We set AVF as the reference throughout the analysis except for the propensity score matching analyses, in which AS was used as the reference. All statistical tests were two-sided, with p value <0.05 considered statistically significant. All analyses were performed using STATA version 17 (STATA Corporation, College Station, TX, USA).

Patient Characteristics

The flowchart of patient recruitment is presented in online supplementary Figure 2. There were 226,999 patients aged ≥20 years receiving hemodialysis with a dialysis vintage ≥6 months at the end of 2006. After excluding 43,509 patients, mainly due to missing information regarding vascular access, a total of 183,490 patients were included in the main analysis. Baseline characteristics, such as age, sex, BMI, dialysis vintage, hemodialysis duration, blood flow, type of vascular access at the hemodialysis initiation, cause of ESKD, laboratory data, and within-facility proportion of patients with AVF, did not differ significantly between patients included in the analysis and those excluded (online suppl. Table 1). However, excluded patients were more likely to have a higher prevalence of myocardial infarction and cerebral infarction.

Of the 183,490 patients included in the analysis, the proportion of those with each type of vascular access was as follows: AVF, n = 166,493 (90.7%); AVG, n = 12,667 (6.9%); AS, n = 3,661 (2.0%); and TC-CVC, n = 669 (0.4%). The baseline characteristics of the study participants by type of vascular access are presented in Table 1. Compared to patients with AVF, those with AVG or AS were more likely to be older and female, and more likely to have a history of cerebral infarction. Moreover, patients with AS had a higher prevalence of myocardial infarction than those with AVF. Compared to patients with AVF, those with TC-CVC were more likely to be older and have more comorbidities, such as cardiovascular disease. Thus, the levels of hemoglobin, albumin, and phosphate were lower in patients with TC-CVC than in those with AVF.

Table 1.

Baseline characteristics according to the vascular access type

CharacteristicsAVFAVGASTC-CVCp value
Missing value, n (%)n = 166,493 (90.7%)n = 12,667 (6.9%)n = 3,661 (2.0%)n = 669 (0.4%)
Age, years 64.0 (12.6) 66.9 (11.7) 68.6 (11.8) 71.6 (12.5) <0.001 
Female, n (%) 12 (<0.1) 62,746 (37.7) 6,718 (53.0) 1,770 (48.3) 426 (63.7) <0.001 
BMI, kg/m2 34,780 (19.0) 22.0 (4.0) 22.0 (3.9) 21.3 (3.8) 20.7 (3.7) <0.001 
Dialysis vintage, years 61 (<0.1) 5 [2–10] 6 [3–12] 6 [2–13] 4 [1–10] <0.001 
Hemodialysis duration, hours/week 1,197 (0.7) 11.6 (1.8) 11.6 (1.7) 11.4 (2.1) 10.8 (2.2) <0.001 
Blood flow, mL/min 2,651 (1.4) 198 (32) 192 (32) 185 (34) 169 (30) <0.001 
Type of vascular access at the initiation of hemodialysis 176,350 (96.1)     <0.001 
 AVF  5,008 (76.2) 176 (44.7) 49 (38.3) 11 (25.6)  
 AVG  30 (0.5) 101 (25.6) 2 (1.6)  
 AS  27 (0.4) 6 (1.5) 27 (21.1) 1 (2.3)  
 TC-CVC  69 (1.0) 2 (0.5) 1 (0.8) 12 (27.9)  
 NT-CVC  1,441 (21.9) 109 (27.7) 49 (38.3) 19 (44.2)  
Cause of ESKD, n (%) 237 (0.1)     <0.001 
 Glomerulonephritis  67,579 (40.6) 4,960 (39.2) 1,400 (38.3) 221 (33.1)  
 Diabetes  50,041 (30.1) 3,994 (31.6) 1,129 (30.9) 250 (37.4)  
 Nephrosclerosis  10,655 (6.4) 760 (6.0) 298 (8.1) 45 (6.7)  
 Polycystic kidney disease  5,544 (3.3) 447 (3.5) 133 (3.6) 13 (1.9)  
 Others  21,405 (12.9) 1,585 (12.5) 433 (11.8) 90 (13.5)  
 Unknown  11,051 (6.6) 906 (7.2) 265 (7.2) 49 (7.3)  
History, n (%) 
 Myocardial infarction 17,221 (9.4) 9,026 (6.0) 764 (6.6) 350 (10.5) 44 (7.5) <0.001 
 Cerebral infarction 17,522 (9.5) 18,244 (12.1) 1,930 (16.7) 563 (16.9) 167 (28.5) <0.001 
 Cerebral hemorrhage 17,225 (9.4) 5,675 (3.8) 605 (5.2) 158 (4.7) 57 (9.7) <0.001 
 Limb amputation 15,275 (8.3) 3,465 (2.3) 442 (3.8) 169 (5.1) 47 (8.0) <0.001 
Laboratory tests 
 Hemoglobin, g/dL 3,004 (1.6) 10.3 (1.3) 10.2 (1.3) 10.1 (1.4) 9.7 (1.6) <0.001 
 Albumin, g/dL 10,998 (6.0) 3.8 (0.4) 3.7 (0.4) 3.6 (0.4) 3.4 (0.6) <0.001 
 Albumin-adjusted calcium, mg/dL 2,429 (1.3) 9.3 (0.9) 9.4 (0.9) 9.4 (0.9) 9.5 (0.9) <0.001 
 Phosphate, mg/dL 2,399 (1.3) 5.4 (1.5) 5.2 (1.5) 5.2 (1.5) 4.8 (1.7) <0.001 
 Intact PTH, pg/mL 26,106 (14.2) 137 [64–244] 122 [56–230] 129 [58–236] 118 [52–220] <0.001 
 Ferritin, ng/mL 25,374 (13.8) 137 [54–284] 137 [54–279] 127 [50–287] 157 [59–360] 0.005 
 Proportion of AVFs within the four types of vascular access in each facility, % 93.2 [88.5–96.7] 85.5 [76.3–91.2] 88.5 [83.5–92.6] 88.5 [82.7–92.4] <0.001 
CharacteristicsAVFAVGASTC-CVCp value
Missing value, n (%)n = 166,493 (90.7%)n = 12,667 (6.9%)n = 3,661 (2.0%)n = 669 (0.4%)
Age, years 64.0 (12.6) 66.9 (11.7) 68.6 (11.8) 71.6 (12.5) <0.001 
Female, n (%) 12 (<0.1) 62,746 (37.7) 6,718 (53.0) 1,770 (48.3) 426 (63.7) <0.001 
BMI, kg/m2 34,780 (19.0) 22.0 (4.0) 22.0 (3.9) 21.3 (3.8) 20.7 (3.7) <0.001 
Dialysis vintage, years 61 (<0.1) 5 [2–10] 6 [3–12] 6 [2–13] 4 [1–10] <0.001 
Hemodialysis duration, hours/week 1,197 (0.7) 11.6 (1.8) 11.6 (1.7) 11.4 (2.1) 10.8 (2.2) <0.001 
Blood flow, mL/min 2,651 (1.4) 198 (32) 192 (32) 185 (34) 169 (30) <0.001 
Type of vascular access at the initiation of hemodialysis 176,350 (96.1)     <0.001 
 AVF  5,008 (76.2) 176 (44.7) 49 (38.3) 11 (25.6)  
 AVG  30 (0.5) 101 (25.6) 2 (1.6)  
 AS  27 (0.4) 6 (1.5) 27 (21.1) 1 (2.3)  
 TC-CVC  69 (1.0) 2 (0.5) 1 (0.8) 12 (27.9)  
 NT-CVC  1,441 (21.9) 109 (27.7) 49 (38.3) 19 (44.2)  
Cause of ESKD, n (%) 237 (0.1)     <0.001 
 Glomerulonephritis  67,579 (40.6) 4,960 (39.2) 1,400 (38.3) 221 (33.1)  
 Diabetes  50,041 (30.1) 3,994 (31.6) 1,129 (30.9) 250 (37.4)  
 Nephrosclerosis  10,655 (6.4) 760 (6.0) 298 (8.1) 45 (6.7)  
 Polycystic kidney disease  5,544 (3.3) 447 (3.5) 133 (3.6) 13 (1.9)  
 Others  21,405 (12.9) 1,585 (12.5) 433 (11.8) 90 (13.5)  
 Unknown  11,051 (6.6) 906 (7.2) 265 (7.2) 49 (7.3)  
History, n (%) 
 Myocardial infarction 17,221 (9.4) 9,026 (6.0) 764 (6.6) 350 (10.5) 44 (7.5) <0.001 
 Cerebral infarction 17,522 (9.5) 18,244 (12.1) 1,930 (16.7) 563 (16.9) 167 (28.5) <0.001 
 Cerebral hemorrhage 17,225 (9.4) 5,675 (3.8) 605 (5.2) 158 (4.7) 57 (9.7) <0.001 
 Limb amputation 15,275 (8.3) 3,465 (2.3) 442 (3.8) 169 (5.1) 47 (8.0) <0.001 
Laboratory tests 
 Hemoglobin, g/dL 3,004 (1.6) 10.3 (1.3) 10.2 (1.3) 10.1 (1.4) 9.7 (1.6) <0.001 
 Albumin, g/dL 10,998 (6.0) 3.8 (0.4) 3.7 (0.4) 3.6 (0.4) 3.4 (0.6) <0.001 
 Albumin-adjusted calcium, mg/dL 2,429 (1.3) 9.3 (0.9) 9.4 (0.9) 9.4 (0.9) 9.5 (0.9) <0.001 
 Phosphate, mg/dL 2,399 (1.3) 5.4 (1.5) 5.2 (1.5) 5.2 (1.5) 4.8 (1.7) <0.001 
 Intact PTH, pg/mL 26,106 (14.2) 137 [64–244] 122 [56–230] 129 [58–236] 118 [52–220] <0.001 
 Ferritin, ng/mL 25,374 (13.8) 137 [54–284] 137 [54–279] 127 [50–287] 157 [59–360] 0.005 
 Proportion of AVFs within the four types of vascular access in each facility, % 93.2 [88.5–96.7] 85.5 [76.3–91.2] 88.5 [83.5–92.6] 88.5 [82.7–92.4] <0.001 

Values are presented as means (standard deviation) or medians [interquartile range] unless otherwise indicated.

AS, arterial superficialization; AVF, arteriovenous fistula; AVG, arteriovenous graft; BMI, body mass index; ESKD, end-stage kidney disease; NT-CVC, non-tunneled, non-cuffed central venous catheter; PTH, parathyroid hormone; TC-CVC, tunneled and cuffed central venous catheter.

Association between Vascular Access Type and Mortality

During the 1-year follow-up, 13,798 patients died (4,598 deaths from cardiovascular diseases and 1,657 deaths from infectious diseases). The Kaplan-Meier survival curves for all-cause and cause-specific mortality by the type of vascular access are presented in Fig. 1a–c. In the multivariable Cox model, patients with AVF had the lowest risk of all-cause mortality, indicating incrementally higher risk in patients with an AVG, AS, and TC-CVC compared to those with AVF: the adjusted HRs [95% confidence interval] were 1.30 (1.20–1.41), 1.56 (1.39–1.76), and 2.15 (1.77–2.61) for patients with an AVG, AS, and TC-CVC, respectively (Table 2 and online suppl. Table 2). Moreover, AVF usage was associated with the lowest risk of infection-related and cardiovascular mortality after adjustment for demographic and clinical factors (Table 2; online suppl. Table 2). The all-cause and cause-specific mortality of patients with AS was intermediate between those with AVG and those with TC-CVC.

Fig. 1.

Kaplan-Meier survival curves according to vascular access type. a All-cause mortality. b Infection-related mortality. c Cardiovascular mortality. AVF, arteriovenous fistula; AVG, arteriovenous graft; AS, arterial superficialization; TC-CVC, tunneled and cuffed central venous catheter.

Fig. 1.

Kaplan-Meier survival curves according to vascular access type. a All-cause mortality. b Infection-related mortality. c Cardiovascular mortality. AVF, arteriovenous fistula; AVG, arteriovenous graft; AS, arterial superficialization; TC-CVC, tunneled and cuffed central venous catheter.

Close modal
Table 2.

All-cause and cause-specific mortality according to the vascular access type

OutcomesUnadjusted HR (95% CI), n = 183,490p valueAdjusted HR (95% CI), n = 104,913p valueAdjusted SHR (95% CI), n = 104,913p value
All-cause mortality 
 AVF ref – ref – – – 
 AVG 1.61 (1.53–1.71) <0.001 1.30 (1.20–1.41) <0.001 – – 
 AS 2.32 (2.13–2.52) <0.001 1.56 (1.39–1.76) <0.001 – – 
 TC-CVC 5.64 (4.94–6.43) <0.001 2.15 (1.77–2.61) <0.001 – – 
Infection-related mortality 
 AVF ref – ref – ref – 
 AVG 1.99 (1.72–2.31) <0.001 1.47 (1.18–1.84) 0.001 1.47 (1.18–1.83) 0.001 
 AS 3.09 (2.49–3.83) <0.001 1.75 (1.29–2.36) <0.001 1.66 (1.22–2.27) 0.001 
 TC-CVC 7.42 (5.28–10.43) <0.001 2.75 (1.75–4.33) <0.001 2.42 (1.48–3.96) <0.001 
Cardiovascular mortality 
 AVF ref – ref – ref – 
 AVG 1.56 (1.41–1.72) <0.001 1.35 (1.17–1.55) <0.001 1.35 (1.17–1.55) <0.001 
 AS 2.37 (2.06–2.74) <0.001 1.50 (1.22–1.85) <0.001 1.44 (1.18–1.77) <0.001 
 TC-CVC 6.02 (4.82–7.51) <0.001 2.58 (1.87–3.56) <0.001 2.37 (1.69–3.31) <0.001 
OutcomesUnadjusted HR (95% CI), n = 183,490p valueAdjusted HR (95% CI), n = 104,913p valueAdjusted SHR (95% CI), n = 104,913p value
All-cause mortality 
 AVF ref – ref – – – 
 AVG 1.61 (1.53–1.71) <0.001 1.30 (1.20–1.41) <0.001 – – 
 AS 2.32 (2.13–2.52) <0.001 1.56 (1.39–1.76) <0.001 – – 
 TC-CVC 5.64 (4.94–6.43) <0.001 2.15 (1.77–2.61) <0.001 – – 
Infection-related mortality 
 AVF ref – ref – ref – 
 AVG 1.99 (1.72–2.31) <0.001 1.47 (1.18–1.84) 0.001 1.47 (1.18–1.83) 0.001 
 AS 3.09 (2.49–3.83) <0.001 1.75 (1.29–2.36) <0.001 1.66 (1.22–2.27) 0.001 
 TC-CVC 7.42 (5.28–10.43) <0.001 2.75 (1.75–4.33) <0.001 2.42 (1.48–3.96) <0.001 
Cardiovascular mortality 
 AVF ref – ref – ref – 
 AVG 1.56 (1.41–1.72) <0.001 1.35 (1.17–1.55) <0.001 1.35 (1.17–1.55) <0.001 
 AS 2.37 (2.06–2.74) <0.001 1.50 (1.22–1.85) <0.001 1.44 (1.18–1.77) <0.001 
 TC-CVC 6.02 (4.82–7.51) <0.001 2.58 (1.87–3.56) <0.001 2.37 (1.69–3.31) <0.001 

Both Cox proportional hazard models and the Fine and Gray models were adjusted for age, sex, body mass index, cause of end-stage kidney disease, dialysis vintage, hemodialysis duration, history of myocardial infarction, cerebral infarction, cerebral hemorrhage and amputation of limbs, laboratory data (albumin, hemoglobin, albumin-adjusted calcium, phosphate, intact PTH, and ferritin), and proportion of AVFs within the four types of vascular access in each facility.

AS, arterial superficialization; AVF, arteriovenous fistula; AVG, arteriovenous graft; CI, confidence interval; ref, reference; HR, hazard ratio; SHR, subdistribution hazard ratio; TC-CVC, tunneled and cuffed central venous catheter.

In subgroup analyses, the association of access type with all-cause mortality was significantly modified by the presence of cerebral infarction and hemorrhage (p value for interaction = 0.040 and 0.049, respectively; Fig. 2a, b). However, we found no significant effect modification by the history of myocardial infarction and limb amputation on these associations (p value for interaction = 0.55 and 0.58, respectively). No other significant interactions were identified in the subgroup analyses.

Fig. 2.

Subgroup analyses of all-cause mortality according to the baseline characteristics. a Subgroup analyses according to age, sex, dialysis vintage, and cause of ESKD. b Subgroup analyses according to vascular access at hemodialysis and study initiation, myocardial infarction, cerebral infarction, cerebral hemorrhage, and limb amputation. Cox proportional hazard models were adjusted for age, sex, body mass index, cause of ESKD, dialysis vintage, and duration and frequency of dialysis, comorbidities (history of myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation of limbs), laboratory data (albumin, hemoglobin, albumin-adjusted calcium, phosphate, intact PTH, and ferritin), and proportion of AVFs within the four types of vascular access in each facility. Bars present 95% CIs. AS, arterial superficialization; AVF, arteriovenous fistula; AVG, arteriovenous graft; CI, confidence interval; ESKD, end-stage kidney disease; HR, hazard ratio; TC-CVC, tunneled and cuffed central venous catheter.

Fig. 2.

Subgroup analyses of all-cause mortality according to the baseline characteristics. a Subgroup analyses according to age, sex, dialysis vintage, and cause of ESKD. b Subgroup analyses according to vascular access at hemodialysis and study initiation, myocardial infarction, cerebral infarction, cerebral hemorrhage, and limb amputation. Cox proportional hazard models were adjusted for age, sex, body mass index, cause of ESKD, dialysis vintage, and duration and frequency of dialysis, comorbidities (history of myocardial infarction, cerebral infarction, cerebral hemorrhage, and amputation of limbs), laboratory data (albumin, hemoglobin, albumin-adjusted calcium, phosphate, intact PTH, and ferritin), and proportion of AVFs within the four types of vascular access in each facility. Bars present 95% CIs. AS, arterial superficialization; AVF, arteriovenous fistula; AVG, arteriovenous graft; CI, confidence interval; ESKD, end-stage kidney disease; HR, hazard ratio; TC-CVC, tunneled and cuffed central venous catheter.

Close modal

The results remained unchanged in the two sensitivity analyses. When missing values were imputed by multiple imputation, the use of AVG, AS, and TC-CVC was associated with an incrementally higher risk of all-cause and cause-specific mortality compared with AVF usage (online suppl. Table 3). In the first propensity score-matched cohort of patients with AS and TC-CVC, the baseline characteristics were well balanced between the two groups (online suppl. Table 4). Patients using TC-CVC had an increased risk of mortality from all causes and cardiovascular diseases compared with those using AS (online suppl. Table 5). In the second propensity score-matched cohort of patients with AS and AVG, patients using AVG had a lower risk of all-cause mortality than those using AS (data not shown).

This nationwide cohort study, including 183,490 patients undergoing maintenance hemodialysis in Japan, compared the risk of mortality based on vascular access types. We found that AVF usage was associated with the lowest risk of all-cause mortality compared with the usage of AVG, AS, and TC-CVC, during the 1-year follow-up period. The all-cause mortality of patients using AS was intermediate between those with AVG and those with TC-CVC. Moreover, sensitivity analyses using multiple imputation and propensity score matching yielded similar results.

To our knowledge, this is the first study comparing mortality with the four types of vascular access: AVF, AVG, AS, and TC-CVC. We showed that AS usage was associated with a lower risk of all-cause and cardiovascular mortality compared with the use of TC-CVC. To date, only two observational studies have examined the association of AS use with all-cause mortality. One study including 60 Japanese patients with heart failure reported that AS usage was associated with a reduced risk of death compared with TC-CVC usage (adjusted HR, 0.30; 95% confidence interval, 0.14–0.65) [25]. However, this study could not adjust for important confounders because of the small sample size. In another study conducted in Japan, there was no significant difference in mortality between patients with AS and those with TC-CVC [26]. Although the JSDT guideline recommends the usage of either AS or TC-CVC for patients in whom AVF or AVG is not suitable [9], our results suggested the superiority of AS to TC-CVC for an alternative access modality in those patients. Considering that AS was reported to have better patency than TC-CVC [26], our results may be useful for selecting the optimal type of vascular access, especially in patients who have difficulty creating or maintaining an AVF or AVG.

There are conflicting results regarding the most optimal vascular access, especially in elderly patients. Several recent observational studies have shown that AVF usage was associated with lower mortality compared with AVG usage even in elderly patients [36, 39], which was consistent with the results of our subgroup analyses. It has been proposed that vascular access-related complications, such as bacteremia, contribute to the increased risk of death among such patients [4]. In addition, a few studies have suggested that artificial materials (AVG and TC-CVC) contribute to cardiovascular diseases and mortality associated with chronic inflammation [40, 41]. However, recent cohort studies in the USA have reported that the majority of deaths were not directly caused by access-related complications [18, 21], suggesting that much of the excess mortality observed in patients with AVG may be attributed to unmeasured confounding factors or patient selection [42]. TC-CVC appears to be used more often in patients with poor health status due to the high burden of comorbidities [18, 43, 44], consistent with this study. To our knowledge, only one pilot randomized controlled trial (RCT) compared clinical outcomes between an approach of AVF and AVG creation in elderly patients who started hemodialysis using TC-CVC [45]. During the follow-up period, access-related infection and CVC-related bacteremia were higher in the AVG than in the AVF group (23% vs. 13% and 31% vs. 13%, respectively) [45]. Larger RCTs with longer follow-up periods are necessary to analyze the clinical differences between the first creation of AVF and AVG, especially in elderly patients.

Moreover, a key finding of our study is that in subgroup analyses, only prior cerebral infarction and cerebral hemorrhage were significant effect modifiers in the association between the type of vascular access and all-cause mortality. The advantage of AVF over other vascular access was attenuated in patients with cerebral hemorrhage, but not in patients with cerebral infarction. Although this discrepancy in results cannot be concluded from the present data, this could be attributed to two possible reasons. First, it could be the high mortality rate after cerebrovascular events, including cerebral hemorrhage [46]. Cerebrovascular disease was reported to pose a high risk of all-cause mortality in the acute and chronic phases [47], in consistency with our findings (adjusted HRs of 1.32 [1.20–1.45] in patients with chronic phase of cerebral hemorrhage). The other reason could be the physical disability due to cerebrovascular disease. Deaths after cerebrovascular disease have been reported mainly due to cardiac disease and infections [47]; however, withdrawal from dialysis was the second leading cause of death [47]. Physical disability is associated with death from dialysis withdrawal [48]. Although data on activities of daily living (ADLs) could not be collected, a previous study in Japan reported that 53% of patients with cerebrovascular disease undergoing hemodialysis had impaired ADLs at 4 weeks after admission [49], and regardless of the cause, physical disability was a significant predictor of 1-year mortality in dialysis patients [32]. Therefore, it may be difficult to detect an association between the type of vascular access and mortality, as cerebral hemorrhage has a significant impact on mortality, including the subsequent decline in ADLs. The results of this study suggest that the creation and maintenance of an AVF are not necessarily advantageous in patients with cerebral hemorrhage, supporting a patient-centered, individualized approach based on a shared decision-making process [50].

This study has several limitations. First, despite adjusting for a large number of patient characteristics that were measured, there was a potential for residual confounders and selection bias. Thus, our study could overestimate the impact of AVF usage. Second, we cannot exclude that lead-time bias would affect the observed mortality differences. Although we excluded patients with a dialysis vintage <6 months to mitigate this bias, further studies, especially RCTs, are necessary to clarify the association between the type of vascular access and mortality. Third, an independent committee could not adjudicate the cause of death. However, all causes of death have been diagnosed and reported by on-site physicians, primarily based on the patient’s medical records. Fourth, we did not have information regarding vascular access changes during the study period. Although this might have affected our results, contamination between the groups would have biased our results toward the null [13]. Fifth, we were unable to determine the vascular access-related complications or the actual mortality rates resulting from these complications. However, a recent multicenter study in Japan reported that the most common cause of hospitalization was related to complications of vascular access among patients undergoing maintenance hemodialysis (29.5%) [51]. In particular, with regard to AS complications, 40 (17%) out of 233 Japanese patients undergoing hemodialysis who newly underwent brachial artery superficialization were reported to have developed complications during a mean follow-up period of 19 months. Impaired wound healing (5.6%) was the most common complication, followed by large aneurysm (2.6%), arterial thrombosis (2.1%), hand ischemia (2.1%), skin infection (2.1%), and arterial stenosis (1.7%) [52]. Sixth, we did not have data concerning the site of AS. However, it has been reported that more than 90% of Japanese patients on hemodialysis use the brachial artery as their SA [9]. Finally, considering that practice patterns regarding vascular access, such as blood flow, vary widely from country to country [22], the results of this study based on the JRDR may not be generalizable to patients undergoing maintenance hemodialysis in other countries. Despite these limitations, our study included a large sample size that is representative of the entire population of patients undergoing maintenance hemodialysis in Japan.

This nationwide cohort study in Japan suggested that the usage of AVF may be associated with improved overall survival compared with that of the other three types of vascular access. This study also suggested that AS usage may be associated with better survival rates compared to those of TC-CVC among patients for whom AVF or AVG is not suitable. Further research is needed to identify patients who may benefit from vascular access other than AVF, especially in patients with comorbidities undergoing hemodialysis, which will help them choose the optimal vascular access through a shared decision-making process with their clinicians.

We thank the Committee of the Renal Data Registry of the JSDT for permitting us to use their data. The opinions reflected in this manuscript are those of the authors alone and do not reflect the official position of the JSDT. We would like to thank Editage (www.editage.com) for English language editing.

The study protocol was approved by the Ethics Committee of the JSDT (approval number 28), and informed consent was waived. The study was conducted according to the principles of the Declaration of Helsinki.

N.F. has received personal fees from Chugai Pharmaceutical, Kissei Pharmaceutical, Kyowa Kirin, Ono Pharmaceutical, and Sanwa Kagaku Kenkyusho. T.H. has received personal fees from Chugai Pharmaceutical, Kissei Pharmaceutical, Kyowa Kirin, Ono Pharmaceutical, and Sanwa Kagaku Kenkyusho; and grants from Chugai Pharmaceutical, Kissei Pharmaceutical, Kyowa Kirin, and Ono Pharmaceutical. I.M. has received personal fees from Chugai Pharmaceutical, Kyowa Kirin, and Ono Pharmaceutical. No other authors have no conflicts of interest to declare.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

M.M., N.F., E.K., K.K., A.W., T.H., and I.M. conceived and designed the research and interpreted the results. A.W. performed data acquisition. M.M. and N.F. analyzed the data. M.M. drafted the manuscript. T.H. and I.M. supervised the study. All authors reviewed and approved the final version of the manuscript.

The datasets generated during and/or analyzed during the current study cannot be made publicly available since they are owned by the JSDT. Further inquiries can be directed to the corresponding author.

1.
GBD Chronic Kidney Disease Collaboration
.
Global, regional, and national burden of chronic kidney disease, 1990-2017: a systematic analysis for the global burden of disease study 2017
.
Lancet
.
2020
;
395
(
10225
):
709
33
.
2.
Liyanage
T
,
Ninomiya
T
,
Jha
V
,
Neal
B
,
Patrice
HM
,
Okpechi
I
.
Worldwide access to treatment for end-stage kidney disease: a systematic review
.
Lancet
.
2015
;
385
(
9981
):
1975
82
.
3.
Johansen
KL
,
Chertow
GM
,
Foley
RN
,
Gilbertson
DT
,
Herzog
CA
,
Ishani
A
.
US renal data system 2020 annual data report: epidemiology of kidney disease in the United States
.
Am J Kidney Dis
.
2021
77
4 Suppl 1
A7
A8
.
4.
Pisoni
RL
,
Arrington
CJ
,
Albert
JM
,
Ethier
J
,
Kimata
N
,
Krishnan
M
.
Facility hemodialysis vascular access use and mortality in countries participating in DOPPS: an instrumental variable analysis
.
Am J Kidney Dis
.
2009
;
53
(
3
):
475
91
.
5.
Lok
CE
,
Foley
R
.
Vascular access morbidity and mortality: trends of the last decade
.
Clin J Am Soc Nephrol
.
2013
;
8
(
7
):
1213
9
.
6.
Thamer
M
,
Lee
TC
,
Wasse
H
,
Glickman
MH
,
Qian
J
,
Gottlieb
D
.
Medicare costs associated with arteriovenous fistulas among US hemodialysis patients
.
Am J Kidney Dis
.
2018
;
72
(
1
):
10
8
.
7.
Robinson
BM
,
Akizawa
T
,
Jager
KJ
,
Kerr
PG
,
Saran
R
,
Pisoni
RL
.
Factors affecting outcomes in patients reaching end-stage kidney disease worldwide: differences in access to renal replacement therapy, modality use, and haemodialysis practices
.
Lancet
.
2016
;
388
(
10041
):
294
306
.
8.
Lawson
JH
,
Niklason
LE
,
Roy-Chaudhury
P
.
Challenges and novel therapies for vascular access in haemodialysis
.
Nat Rev Nephrol
.
2020
;
16
(
10
):
586
602
.
9.
Kukita
K
,
Ohira
S
,
Amano
I
,
Naito
H
,
Azuma
N
,
Ikeda
K
.
2011 update Japanese society for dialysis therapy guidelines of vascular access construction and repair for chronic hemodialysis
.
Ther Apher Dial
.
2015
19
Suppl 1
1
39
.
10.
Lok
CE
,
Huber
TS
,
Lee
T
,
Shenoy
S
,
Yevzlin
AS
,
Abreo
K
.
KDOQI clinical practice guideline for vascular access: 2019 update
.
Am J Kidney Dis
.
2020
75
4 Suppl 2
S1
S164
.
11.
Polkinghorne
KR
,
McDonald
SP
,
Atkins
RC
,
Kerr
PG
.
Vascular access and all-cause mortality: a propensity score analysis
.
J Am Soc Nephrol
.
2004
;
15
(
2
):
477
86
.
12.
Bradbury
BD
,
Chen
F
,
Furniss
A
,
Pisoni
RL
,
Keen
M
,
Mapes
D
.
Conversion of vascular access type among incident hemodialysis patients: description and association with mortality
.
Am J Kidney Dis
.
2009
;
53
(
5
):
804
14
.
13.
Lacson
E
Jr
,
Wang
W
,
Lazarus
JM
,
Hakim
RM
.
Change in vascular access and mortality in maintenance hemodialysis patients
.
Am J Kidney Dis
.
2009
;
54
(
5
):
912
21
.
14.
Lacson
E
Jr
,
Wang
W
,
Lazarus
JM
,
Hakim
RM
.
Change in vascular access and hospitalization risk in long-term hemodialysis patients
.
Clin J Am Soc Nephrol
.
2010
;
5
(
11
):
1996
2003
.
15.
Murea
M
,
James
KM
,
Russell
GB
,
Byrum
GV
3rd
,
Yates
JE
,
Tuttle
NS
.
Risk of catheter-related bloodstream infection in elderly patients on hemodialysis
.
Clin J Am Soc Nephrol
.
2014
;
9
(
4
):
764
70
.
16.
Shingarev
R
,
Barker-Finkel
J
,
Allon
M
.
Natural history of tunneled dialysis catheters placed for hemodialysis initiation
.
J Vasc Interv Radiol
.
2013
;
24
(
9
):
1289
94
.
17.
Ravani
P
,
Palmer
SC
,
Oliver
MJ
,
Quinn
RR
,
MacRae
JM
,
Tai
DJ
.
Associations between hemodialysis access type and clinical outcomes: a systematic review
.
J Am Soc Nephrol
.
2013
;
24
(
3
):
465
73
.
18.
DeSilva
RN
,
Patibandla
BK
,
Vin
Y
,
Narra
A
,
Chawla
V
,
Brown
RS
.
Fistula first is not always the best strategy for the elderly
.
J Am Soc Nephrol
.
2013
;
24
(
8
):
1297
304
.
19.
Brown
RS
,
Patibandla
BK
,
Goldfarb-Rumyantzev
AS
.
The survival benefit of “Fistula First, Catheter Last” in hemodialysis is primarily due to patient factors
.
J Am Soc Nephrol
.
2017
;
28
(
2
):
645
52
.
20.
Ravani
P
,
Quinn
R
,
Oliver
M
,
Robinson
B
,
Pisoni
R
,
Pannu
N
.
Examining the association between hemodialysis access type and mortality: the role of access complications
.
Clin J Am Soc Nephrol
.
2017
;
12
(
6
):
955
64
.
21.
Lyu
B
,
Chan
MR
,
Yevzlin
AS
,
Gardezi
A
,
Astor
BC
.
Arteriovenous access type and risk of mortality, hospitalization, and sepsis among elderly hemodialysis patients: a target trial emulation approach
.
Am J Kidney Dis
.
2022
;
79
(
1
):
69
78
.
22.
Pisoni
RL
,
Zepel
L
,
Port
FK
,
Robinson
BM
.
Trends in US vascular access use, patient preferences, and related practices: an update from the US DOPPS practice monitor with international comparisons
.
Am J Kidney Dis
.
2015
;
65
(
6
):
905
15
.
23.
Pisoni
RL
,
Zepel
L
,
Fluck
R
,
Lok
CE
,
Kawanishi
H
,
Süleymanlar
G
.
International differences in the location and use of arteriovenous accesses created for hemodialysis: results from the dialysis outcomes and practice patterns study (DOPPS)
.
Am J Kidney Dis
.
2018
;
71
(
4
):
469
78
.
24.
Pisoni
RL
,
Zepel
L
,
Zhao
J
,
Burke
S
,
Lok
CE
,
Woodside
KJ
.
International comparisons of native arteriovenous fistula patency and time to becoming catheter-free: findings from the dialysis outcomes and practice patterns study (DOPPS)
.
Am J Kidney Dis
.
2021
;
77
(
2
):
245
54
.
25.
Nakagawa
K
,
Yamada
S
,
Matsukuma
Y
,
Nakano
T
,
Mitsuiki
K
.
Survival comparison between superficialization of the brachial artery and tunneled central venous catheter placement in hemodialysis patients with heart failure: a retrospective study
.
Ther Apher Dial
.
2020
;
24
(
4
):
408
15
.
26.
Soma
Y
,
Murakami
M
,
Nakatani
E
,
Sato
Y
,
Tanaka
S
,
Mori
K
.
Brachial artery transposition versus catheters as tertiary vascular access for maintenance hemodialysis: a single-center retrospective study
.
Sci Rep
.
2022
;
12
(
1
):
306
.
27.
Rao
NN
,
Stokes
MB
,
Rajwani
A
,
Ullah
S
,
Williams
K
,
King
D
.
Effects of arteriovenous fistula ligation on cardiac structure and function in kidney transplant recipients
.
Circulation
.
2019
;
139
(
25
):
2809
18
.
28.
Nakai
S
,
Masakane
I
,
Akiba
T
,
Shigematsu
T
,
Yamagata
K
,
Watanabe
Y
.
Overview of regular dialysis treatment in Japan as of 31 December 2006
.
Ther Apher Dial
.
2008
;
12
(
6
):
428
56
.
29.
von Elm
E
,
Altman
DG
,
Egger
M
,
Pocock
SJ
,
Gøtzsche
PC
,
Vandenbroucke
JP
.
Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies
.
BMJ
.
2007
;
335
(
7624
):
806
8
.
30.
Ravani
P
,
Gillespie
BW
,
Quinn
RR
,
MacRae
J
,
Manns
B
,
Mendelssohn
D
.
Temporal risk profile for infectious and noninfectious complications of hemodialysis access
.
J Am Soc Nephrol
.
2013
;
24
(
10
):
1668
77
.
31.
Nakai
S
,
Masakane
I
,
Shigematsu
T
,
Hamano
T
,
Yamagata
K
,
Watanabe
Y
.
An overview of regular dialysis treatment in Japan (as of 31 December 2007)
.
Ther Apher Dial
.
2009
;
13
(
6
):
457
504
.
32.
Wakasugi
M
,
Kazama
JJ
,
Wada
A
,
Hamano
T
,
Masakane
I
,
Narita
I
.
Functional impairment attenuates the association between high serum phosphate and mortality in dialysis patients: a nationwide cohort study
.
Nephrol Dial Transplant
.
2019
;
34
(
7
):
1207
16
.
33.
Karaboyas
A
,
Morgenstern
H
,
Pisoni
RL
,
Zee
J
,
Vanholder
R
,
Jacobson
SH
.
Association between serum ferritin and mortality: findings from the USA, Japan and European dialysis outcomes and practice patterns study
.
Nephrol Dial Transplant
.
2018
;
33
(
12
):
2234
44
.
34.
Noordzij
M
,
Leffondré
K
,
van Stralen
KJ
,
Zoccali
C
,
Dekker
FW
,
Jager
KJ
.
When do we need competing risks methods for survival analysis in nephrology
.
Nephrol Dial Transplant
.
2013
;
28
(
11
):
2670
7
.
35.
Austin
PC
.
An introduction to propensity score methods for reducing the effects of confounding in observational studies
.
Multivariate Behav Res
.
2011
;
46
(
3
):
399
424
.
36.
Lee
T
,
Thamer
M
,
Zhang
Q
,
Zhang
Y
,
Allon
M
.
Vascular access type and clinical outcomes among elderly patients on hemodialysis
.
Clin J Am Soc Nephrol
.
2017
;
12
(
11
):
1823
30
.
37.
Saleh
T
,
Sumida
K
,
Molnar
MZ
,
Potukuchi
PK
,
Thomas
F
,
Lu
JL
.
Effect of age on the association of vascular access type with mortality in a cohort of incident end-stage renal disease patients
.
Nephron
.
2017
;
137
(
1
):
57
63
.
38.
Jhee
JH
,
Hwang
SD
,
Song
JH
,
Lee
SW
.
The Impact of comorbidity burden on the association between vascular access type and clinical outcomes among elderly patients undergoing hemodialysis
.
Sci Rep
.
2019
;
9
(
1
):
18156
.
39.
Lyu
B
,
Chan
MR
,
Yevzlin
AS
,
Astor
BC
.
Catheter dependence after arteriovenous fistula or graft placement among elderly patients on hemodialysis
.
Am J Kidney Dis
.
2021
;
78
(
3
):
399
408.e1
.
40.
Wasse
H
,
Cardarelli
F
,
De Staercke
C
,
Hooper
WC
,
Long
Q
.
Accumulation of retained nonfunctional arteriovenous grafts correlates with severity of inflammation in asymptomatic ESRD patients
.
Nephrol Dial Transplant
.
2013
;
28
(
4
):
991
7
.
41.
Dukkipati
R
,
Molnar
MZ
,
Park
J
,
Jing
J
,
Kovesdy
CP
,
Kajani
R
.
Association of vascular access type with inflammatory marker levels in maintenance hemodialysis patients
.
Semin Dial
.
2014
;
27
(
4
):
415
23
.
42.
Allon
M
.
Vascular access for hemodialysis patients: new data should guide decision making
.
Clin J Am Soc Nephrol
.
2019
;
14
(
6
):
954
61
.
43.
Grubbs
V
,
Wasse
H
,
Vittinghoff
E
,
Grimes
BA
,
Johansen
KL
.
Health status as a potential mediator of the association between hemodialysis vascular access and mortality
.
Nephrol Dial Transplant
.
2014
;
29
(
4
):
892
8
.
44.
Lee
T
,
Qian
J
,
Thamer
M
,
Allon
M
.
Tradeoffs in vascular access selection in elderly patients initiating hemodialysis with a catheter
.
Am J Kidney Dis
.
2018
;
72
(
4
):
509
18
.
45.
Robinson
T
,
Geary
RL
,
Davis
RP
,
Hurie
JB
,
Williams
TK
,
Velazquez-Ramirez
G
.
Arteriovenous fistula versus graft access strategy in older adults receiving hemodialysis: a pilot randomized trial
.
Kidney Med
.
2021
;
3
(
2
):
248
56.e1
.
46.
Ocak
G
,
Boenink
R
,
Noordzij
M
,
Bos
WJW
,
Vikse
BE
,
Cases
A
.
Trends in mortality due to myocardial infarction, stroke, and pulmonary embolism in patients receiving dialysis
.
JAMA Netw Open
.
2022
;
5
(
4
):
e227624
.
47.
Iseki
K
,
Fukiyama
K
Okawa Dialysis Study OKIDS Group
.
Clinical demographics and long-term prognosis after stroke in patients on chronic haemodialysis. The Okinawa dialysis study (OKIDS) group
.
Nephrol Dial Transplant
.
2000
;
15
(
11
):
1808
13
.
48.
Chen
JHC
,
Brown
MA
,
Jose
M
,
Brennan
F
,
Johnson
DW
,
Roberts
MA
.
Temporal changes and risk factors for death from early withdrawal within 12 months of dialysis initiation-a cohort study
.
Nephrol Dial Transplant
.
2022
;
37
(
4
):
760
9
.
49.
Toyoda
K
,
Fujii
K
,
Fujimi
S
,
Kumai
Y
,
Tsuchimochi
H
,
Ibayashi
S
.
Stroke in patients on maintenance hemodialysis: a 22-year single-center study
.
Am J Kidney Dis
.
2005
;
45
(
6
):
1058
66
.
50.
Viecelli
AK
,
Lok
CE
.
Hemodialysis vascular access in the elderly-getting it right
.
Kidney Int
.
2019
;
95
(
1
):
38
49
.
51.
Shimizu
S
,
Onishi
Y
,
Kabaya
K
,
Wang
J
,
Fukuma
S
,
Morinaga
J
.
Cohort profile: alliance for quality assessment in healthcare-dialysis (AQuAH-D) prospective cohort study of patients on haemodialysis in Japan
.
BMJ Open
.
2022
;
12
(
1
):
e054427
.
52.
Murakami
M
,
Mori
K
,
Hamanoue
S
,
Suemitsu
K
,
Kajiwara
K
,
Miyamoto
M
.
Multicentre study on the efficacy of brachial artery transposition among haemodialysis patients
.
Eur J Vasc Endovasc Surg
.
2021
;
61
(
6
):
998
1006
.