In an aging society, the proportion of elderly patients with end stage kidney disease (ESKD) is rapidly increasing. According to the 2022 annual report of the Korean Renal Data System, the average age of patients undergoing hemodialysis (HD) or peritoneal dialysis (PD) has continued to increase, reaching 67 years in 2022. Moreover, the patients over 65 years old accounted for approximately 60% of ESKD cases [1]. Elderly and non-elderly patients exhibit several important differences in the selection and performance of dialysis. Despite many debates about the preference on dialysis modality in elderly patients, the reality is that elderly patients mostly choose HD in South Korea [2]. Elderly patients encounter several important challenges in maintaining HD, particularly regarding appropriate vascular access [3]. However, the selection of appropriate vascular access for elderly patients with ESKD is complex and demands an individualized approach along with consideration of each individual’s physical condition and life-plan [4]. This review article aims to provide different point of views through a review of existing studies on the selection of vascular access for HD in elderly patients.

Previous studies have reported results on the relationship between vascular access type and mortality in elderly patients undergoing HD (Table 1). Although most studies were observational, the majority of them demonstrated that arteriovenous fistula (AVF) was superior to other vascular access types, even in elderly patients. According to an observational Korean study involving 529 patients on HD, those with arteriovenous graft (AVG) had a fifth rate of vascular maturation failure, but they experienced 2.7 times higher rates of vascular abandonment and mortality [5] compared to those with radio-cephalic AVF. In an observational American study involving 124,421 patients, compared to patients who used AVF, the mortality rate was higher among patients who used AVG (adjusted HR 1.24) or who switched from central venous catheter (CVC) to AVF (adjusted HR 1.36) or AVG (adjusted HR 1.62). In particular, patients who continued to use CVC had the highest mortality rate (adjusted HR 2.23) [6]. However, despite other studies indicating a survival benefit for elderly patients using AVF, the fistula-first strategy has several issues (Table 2) [7].

Table 1 . Studies on the relationship between vascular access types and mortality in elderly patients with hemodialysis.

Year/journal	Subjects (n)	Age	Men	DM	Proportion or aHRs	Initial AVF	CVC to AVF	Initial AVG	CVC to AVG	Persistent CVC
2018/BMC Nephrol. [3]	529	73.6±6.0	61.2%	57.5%	N (%)	190 (35.9)	242 (45.7)	37 (7.0)	60 (11.3)	0
					N (%) of mortality	28 (6.5)		16 (16.5)		0
2019/J. Vasc. Surg. [4]	124,421	82.1	54.8%	86.4%	N (%)	19,173 (15)	29,872 (24)	4,480 (4)	10,712 (9)	59,824 (48)
					aHR for mortality	1	1.36	1.24	1.62	2.23
2019/Sci. Rep. [1]	23,100	73.7±6.0	56.1%	39.1%	N (%)	14,847 (64.3)		3,906 (16.9)		4,347 (18.8)
					aHR for mortality
					Lowest sCCI	1		1.04 (0.67−1.63)		2.81 (1.90−4.16)
					Middle sCCI	1		1.20 (0.88−1.65)		2.54 (1.83−3.53)
					Highest sCCI	1		1.33 (1.00−1.75)		1.43 (1.08−1.90)
2020/NDT. [9]	8,326	84.9±3.3	54%	53%	N (%)	1,511 (18)	3,018 (36)	489 (6)	1,063 (13)	2,245 (27)
					aHR for mortality	0.73	0.79	0.82	0.90	1

aHR, adjusted hazard ratio; AVF, arteriovenous fistula; AVG, arteriovenous graft; CVC, central venous catheter; DM, diabetes mellitus; sCCI, simplified charlson comorbidity index..

Table 2 . Advantages or disadvantages of fistula-first strategy in elderly patients with end stage kidney disease.

Advantages	Disadvantages
Lower risk of infection	Additional procedure prior to use
Lower risk of thrombosis	Failure to use due to patient’s short life expectancy
Lower risk of secondary failure	Higher risk of primary failure

Existing studies reported that the rate of decline in kidney function varies with age. In particular, kidney function was found to decline relatively slowly in elderly patients with advanced chronic kidney disease, where the estimated glomerular filtration rate was less than 25 mL/min [8]. Therefore, elderly patients with ESKD may maintain a premade AVF or AVG for several months or even more than a year without using it. Furthermore, vascular interventions may be required for stenosis or thrombosis of AVF or AVG that have never been used. According to a study by O’Hare et al. [8]. the ratio of intervention before and after AVF use was 0.5:1 in non-elderly patients, whereas it was approximately 5:1 in elderly patients. Although the results were controversial, the majority of previous studies have reported the success rate of AVF was relatively low in elderly patients compared to non-elderly patients [7]. An observational study reported an increased risk of fistula maturation failure in patients aged 65 years and older (OR 2.23) [9]. A meta-analysis of 13 observational studies demonstrated that in elderly patients, both primary and secondary patency of AVF were lower compared to non-elderly patients. Specifically, radio-cephalic AVF had a higher maturation failure rate than other vascular access types [10]. In the elderly patients with short life expectancy, the period required for AVF maturation may exceed patient’s life expectancy. Accordingly, a study suggested that AVF might not be appropriate for elderly patients with short life expectancy [11]. The other study reported that the rate of actual AVF use in elderly patients was lower than the rate of general population undergoing HD [12]. Therefore, the fistula-first strategy appears to be controversial in elderly patients with short life expectancy. In a retrospective study involving 8,356 patients undergoing HD with an average age of 80 years, the mortality rate within a year after HD initiation was highest among patients who maintained CVC. In contrast, the mortality rate was relatively low among those who used AVF or AVG and among those who switched from CVC to AVF. Interestingly, the mortality rate was similar between patients who used AVF and those who switched from CVC to AVF within the first year of dialysis [13]. Based on the results, a rapid transition from initial CVC use to AVF in elderly patients might be considered as an alternative to initial AVF use. However, the appropriateness of such a strategy remains controversial. In many cases, the transition from CVC to AVF is delayed. The study by Lyu et al. [14] showed that the rate of CVC dependence at 3 months after AVF creation was 83%. On the other hand, another study demonstrated a significant increase in the mortality rate after 3 months of CVC use [13]. Therefore, the strategy of a rapid transition from initial CVC use to AVF is not currently considered appropriate.

AVG is usually less preferred compared to AVF, even in elderly patients with ESKD. However, an RCT by Rooijens et al. [15] involving a general population undergoing HD with poor vascular conditions, the patency rate after HD initiation was higher in those who used AVG than in those who used radio-cephalic AVF. Therefore, active AVG selection may be considered more appropriate in patients with poor vascular conditions, including elderly patients.

The recently updated KDOQI guideline for vascular access has emphasized a ‘patient-centered approach’ [16]. According to this approach, specific conditions including old age should not be the sole determinant for selecting vascular access types. Similarly, the fistula-first strategy or the alternative use of AVG or CVC should be also carefully considered depending on the patient’s individual condition. In conclusion, the selection of appropriate vascular access requires various individualized approaches based on comprehensive evaluation.

Existing literatures have suggested that AVF is the preferred vascular access even in elderly patients with ESKD. However, in elderly patients, the fistula-first strategy may not necessarily be applicable. Therefore, individualized approaches, along with considerations of each individual’s physical condition, life expectancy and life-plan are essentially required for elderly patients with ESKD.

The author declares no potential conflict of interest.