CANDIDATES

Complications of Arteriovenous Fistula in Dialysis Patients

H. Salahi, A. Fazelzadeh, A. Mehdizadeh, A. Razmkon, and S.-A. Malek-Hosseini

ABSTRACT
Repeated access to the circulation is essential to perform adequate maintenance hemo-
dialysis (HD). Dysfunction of fistulae is the most common reason for a second intervention
and recurrent hospitalization. The aim of this study was to evaluate the complications of
HD fistulas seeking to evaluate the impact of age, site of arteriovenous fistula (AVF)
(proximal or distal), side (left or right), and history of previous vascular access. We
evaluated the clinical complications in 273 patients from the beginning of the use of the
current access using the history and physical examination obtained at every dialysis session.
We performed further investigations including doppler ultrasound or spiral computed
tomography to confirm the clinical diagnosis. Of our patients, 40% had diabetes mellitus
as the cause of end-stage renal disease. Almost half (49%) the patients dialyzed through
an AVF and 13% with a catheter. One hundred eighty-four cases (67.6%) experienced
complications. Of 145 cases that had elbow AVFs, 103 cases (71%) had complications; of
128 cases with wrist AVFs, 80 cases (62.5%) had complications. There were 115 (62.5%)
complicated cases among 185 patients with left AVFs, and 69 (78%) among 88 patients
with right AVFs. The rate of AVF complications increased with age. The 1-year survival
rate was 94%. We did not observe any significant difference between AVF complications
in patients with diabetes mellitus or hypertension as the underlying cause of renal failure.
Mean cholesterol plasma level did not differ significantly between the patients with and
without AVF complications. Mean hematocrit levels were not significantly different
between the two groups. However, mean EPO weekly dose was significantly higher among
the group of patients with AVF complications. We did find that rate of complications
increased with age (P ⬍ .05). Our results showed that the frequency of complications was
higher among patients with elbow and right-side AVFs, and also among patients with a
history of a previous failed shunt but no significant relationship was observed between
these variables (P ⬎ .05).

R EPEATED ACCESS to the circulation is essential to

perform adequate maintenance hemodialysis (HD).
Complications of vascular access remain the Achilles heel
for many patients with end-stage renal disease (ESRD).
Arteriovenous fistulae (AVF) constructed using native ves-
sels, vascular grafts, and central venous catheters are the
best permanent access, owing to a lower incidence of
stenosis, thrombosis and infection.1 The radiocephalic AVF
of Brescia-Cimina remains the first choice for vascular
access.2 Dysfunction of these fistulas is the most common
reason for a second intervention and for recurrent hospi-
talization, which both increase costs. Therefore, early de-
tection and treatment of complications prevents more
severe conditions and consequent saving of additional costs
and reduction of hospitalization periods. The most frequent
complications are aneurysm formation, vascular steal syn-
drome, venous hypertension, hemorrhage, infection, and
neurologic disorders.3 The aim of this study was to evaluate
complications of HD fistulas to discover the impact of age,
From the Southern Iran Organ Transplant Center, Fars, Iran.
Address reprint requests to Dr H. Salahi, Southern Iran Organ
Transplant Center, PO Box 71455-166, Shiraz, Fars, Iran. E-mail:
ali_ra@yahoo.com

© 2006 by Elsevier Inc. All rights reserved. 0041-1345/06/$–see front matter

360 Park Avenue South, New York, NY 10010-1710 doi:10.1016/j.transproceed.2006.02.066

Transplantation Proceedings, 38, 1261–1264 (2006) 1261

1262
site of AVF (proximal or distal), side (left or right), and
history of previous vascular access.
MATERIAL AND METHODS
We studied 273 patients who underwent hemodialysis in two
centers between 2003 and 2004. The information included age,
lipid profile, diabetes, hypertension, plasma cholesterol, hemato-
crit, erythropoietin (EPO) weekly dose, type and site of AVF,
history of previous access and history of any previous complications
were recorded. All the patients underwent AVF procedures by the
same team in the same hospital. All HD fistulas were native; no
graft materials were used.
The types of AVFs were brachiocephalic, which were con-
structed at the antecubital fossa or radiocephalic, constructed at
the wrist. All patients underwent chronic HD for a mean time of 4.0
hours three times a week. Blood flow rate ranged between 250 and
300 mL/min, whereas a dialysate flow rate of 500 mL/min was
routinely used. Kt/V was evaluated monthly as a marker of dialysis
efficiency. None of the patients had clinical evidence of malignancy,
procoagulant conditions, or took immunosuppressive medications
during the period of observation.
Cardiovascular risk factors such as diabetes mellitus and hyper-
tension were ascertained through personal interview or hospital
records. In particular, study participants were classified as diabetic
or nondiabetic. Hypertension was defined as a blood pressure levels
higher than 140/90 mm Hg or the regular use of at least one
antihypertensive agent. The EPO weekly dose was obtained by
medical record review. The target hematocrit levels were 32% to
34% for all hemodialysis patients. Hematocrit and serum choles-
terol levels were studied monthly by standard laboratory methods.
The mean values of EPO weekly dose, hematocrit, and cholesterol
were calculated for each patient over the whole period of obser-
vation.
We recorded all complications that developed since beginning
the use of the current accesses up to 2004. Access dysfunction was
determined by a variety of clinical and physiologic parameters,
including physical examination, patient biochemistry, urea kinetic
modeling, and raised venous dialysis pressures at every dialysis
session in each dialysis center.
We define complications clinically in Table 1. If any clinical
complications were observed we performed further investigations
including Doppler ultrasound or spiral computed tomography to
confirm the clinical diagnosis. Patients were censored at discontin-
uation of dialysis because of transplantation or death.
The data were analyzed using ␹2 test, Pearson test, or Fisher’s
exact test. P values ⬍ .05 were considered significant. Categorical
variables were expressed as absolute values and percentages.
Table 1. Main Clinical Features of Patients With or Without
Arteriovenous Fistula Complications
With Complication Without Complication
Cases (n) 184 89
Time on dialysis (mo) 72.1 ⫾ 61.2 69.6 ⫾ 67.2
Kt/V 1.42 ⫾ 0.1 1.46 ⫾ 0.2
Diabetic number 58 51
Hypertension number 50 44
Plasma cholesterol 181 ⫾ 40 176 ⫾ 46
Hematocrit % 31.1 ⫾ 0.6 30.6 ⫾ 0.4
EPO weekly dose* 5643 ⫾ 730 4111 ⫾ 419
Abbreviation: EPO, erythropoietin.
*P value ⫽ .029.
SALAHI, FAZELZADEH, MEHDIZADEH ET AL
Table 2. Criteria for AVF Complications
Thrombosis Pain, cyanotic swelling, skin thickness or
ulceration of the involved extremity
Venous hypertension Gradually painfull swelling and
hyperpigmentation of extremity
accompanied by engorgement of
superficial veins
Aneurysm Dilation of vein walls
Vascular steal Coolness, pallor, pain, neuropraxia,
syndrome ischemic gangrene of the graft that get
worst on dialysis
Infection Cellulites: erythema, edema and
tenderness around the graft
Neurologic disorder Occasional tingling, paresthesia, burning
sensation, hotness, pain limitation of
range of motion in joints, flexion
contracture and muscular weakness and
atrophy
Hemorrhage Bloody discharge from the site of fistula
RESULTS
The main clinical features of the patients included in the
study are summarized in Table 1. We first investigated the
presence of classic and novel risk factors for vascular
disease in our patients. Forty percent had diabetes mellitus.
Almost half (49%) of the patients were dialyzed through an
AVF and 13% with a catheter. Average blood flows were
288.5 mL/min. The mean urea reduction ratio and single-
pool Kt/V of the population were 70% and 1.47, respec-
tively. One-year patient survival was 94%. Hypercholester-
olemia is a well known factor in atherosclerosis. However,
mean cholesterol plasma level did not differ significantly
between the patients with and without AVF complications
(see Table 1).
Mean hematocrit levels were not significantly different
between the two groups (Table 1). On the other hand, mean
EPO weekly dose was significantly higher among the groups
of patients with AVF complications (Table 1).
Of the 273 patients followed since inception at both
centers, 184 cases (67.6%) among the study population had
any of the noted AVF complications described in Table 2.
The study of site and side of AVFs, as showed in Table 3,
revealed that 103 cases (71%) of the 145 cases who had
elbow AVFs, experienced complications and 80 cases
(62.5%) of 128 cases with wrist AVFs had complications.
There were 115 (62.5%) complicated cases among 185
patients with left AVFs, and 69 (78%) complicated cases
among 88 patients with right AVFs. Among the 89 patients
with a history of previous failed AVF, 75 cases (84.2%)
experienced new AVF complications. The study of AVF in
proportion to age showed that the rate of AVF complica-
tions increased with age; most complications were seen in
patients older than 50 years (Table 3).
We did not find any significant relation between presence
of hypertension and AVF complications. There was no
significant difference between the primary failure rate, fistula
maturation rate, revision rate or incidence of complications
COMPLICATIONS IN DIALYSIS 1263

between diabetic and nondiabetic patients (Table 1). The

80 (62.5%)

115 (62.2%)

75 (84.2%)

6 (42.8%)
11 (42.3%)
14 (31.1%)
32 (38.5%)

184 (67.6%)
Summation

103 (71%)

69 (78%)

109 (59%)

2 (40%)

43 (43%)
absolute values and percentages of observed AVF compli-
cations is shown in Table 3.

DISCUSSION
Amputation

1 (0.8%)

1 (1.1%)

1 (0.5%)

1 (0.4%)
In 1997, at least $1.2 billion was spent by Centers for
0 (0%)

0 (0%)

0 (0%)

0 (0%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
1 (1%)
Medicare and Medicaid Services on the establishment and
maintenance of vascular angioaccess for the approximately
250,000 Medicare-eligible beneficiaries on maintenance he-
modialysis.4 Moreover, vascular access-related hospitaliza-
1 (0.7%)

1 (1.1%)

1 (1.1%)

1 (0.4%)
Infection

0 (0%)

0 (0%)

0 (0%)

0 (0%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
1 (1%)
tions were the most frequent DRG for Medicare-eligible
Table 3. Absolute Value and Percentages of Any Observed Complication in All 273 Cases Under Study

ESRD patients, accounting for 45% of hospitalizations

from 1995 through 1997.4 Therefore, expenditures for
vascular access are the single greatest categorical expense
Hypertension

18 (12.4%)

11 (12.5%)

12 (13.6%)
7 (5.6%)

14 (7.6%)

13 (7.1%)

1 (3.8%)
3 (6.6%)
7 (8.4%)

25 (9.2%)
10 (10%)
Venous

3 (21.4)

for ESRD care in the United States.

1 (20)

Moreover, vascular access remains a major source of

morbidity and mortality for hemodialysis patients. So, de-
creasing the complications of vascular access can improve
both economy and survival of patients. Several risk factors
Hemorrhage

1 (0.7%)

1 (1.1%)

1 (1.1%)

1 (1.2%)

1 (0.4%)
0 (0%)

0 (0%)

0 (0%)

0 (0%)
0 (0%)
0 (0%)
0 (0%)

0 (0%)

for fistula complications has been identified including fe-

male gender, black race, age above 64 years, and underlying
diabetes.3,5 Studies evaluating the lengths of hospital stay
for access surgery showed that it was prolonged (⬎14 days)
Thrombosis

for elderly patients, usually because of postoperative fever,

4 (3.1%)

2 (1.1%)
2 (2.3%)

4 (2.2%)

1 (2.2%)
2 (2.4%)

4 (1.5%)
0 (0%)

0 (0%)

0 (0%)
0 (0%)
0 (0%)

1 (1%)

repeated femoral catheterization and infection.6

In our study, the rate of complications increased with age
(P ⬍ .05); there was a 35% difference in complicated cases
among patients above 50 years and those under 18 years (8
Vascular Steal

11 (12.5%)

10 (11.4%)

9 (10.8%)

versus 110). No vascular steal syndrome, thrombosis, or

Syndrome

14 (9.7%)
12 (9.4%)

15 (8.1%)

16 (8.7%)

1 (3.8%)
4 (8.8%)

26 (9.5%)
12 (12%)
0 (0%)
0 (0%)

neurologic disorder was observed under the age of 18 years.

The most commonly used AVF site is the wrist, connect-
ing the cephalic vein to the radial artery (Brescia-Cimino
fistula) or the basilic vein to the ulnar artery.2,7 Some
25 (17.2%)
15 (11.7%)

22 (11.9%)
18 (20.5%)

17 (19.3%)
23 (12.5%)

3 (11.5%)
6 (13.3%)
13 (15.6%)

40 (14.7%)

studies have reported that one year patency rate for these
Neurologic
Disorder

18 (18%)
0 (0%)
0 (0%)

fistulas are as high as 82%.8

In our study, the rate of 1-year patency of AVFs was
32.4%. The results from our present study compared favor-
ably with those achieved by interventional radiologists,
32 (38.55%)

particularly with regard to primary patency. Reports of

62 (33.5%)
24 (27.3%)

28 (31.7%)
58 (31.5%)

12 (26.6%)

86 (31.5%)
Aneurysm

45 (31%)
41 (32%)

1 (20%)

6 (23%)

32 (32%)
3 (21.4)

primary patency after percutaneous angioplasty of stenoses

in native AVFs vary widely (16% to 60% 1-year primary
patency)9 –11 with the best reports of 62% 1-year primary
patency after percutaneous angioplasty of juxta-anastomotic
venous stenoses in Bresica fistulas.9
273 (100%)
Total Shunt

Our results showed that the frequency of complications

145
128

185
88

89
184

5
14
26
45
83
100

was higher among patients with elbow and right-side AVFs

and also those with a history of a previous failed shunt, but
no significant relationship was observed between these
variables (P ⬎ .05).
Absolute Value (Percent)
Previous access history

In our study “thrombosis,” despite the low incidence, was

Proximal (elbow)

a serious fatal complication of AVF; all patients with this

complication died during our study. This finding shows that
Distal (wrist)
Side of AVF

there must be better management of complicated AVFs to

Site of AVF

Negative
Positive

decrease the mortality rate.

15–18
18–25
25–40
40–60
Right

Age (y)
⬍15

⬎60
Left

The 1-year patient survival rate was 94%. But this

survival rate does not reflect the true survival rate of dialysis
1264
patients, because we estimated it in a small cohort of
patients who underwent dialysis in two big centers and it did
not include results at the small dialysis centers that have
little facilities and therefore show greater mortality rates.
Some studies showed that the most frequent finding in
patients with AVF is stenosis at the venous anastomosis.
Alteration in coagulation may play a role due to increased
hematocrit from transfusions or from erythropoithin therapy.12
In our study 86 cases of “aneurysm formation” provided
the most common complication. There was neurologic
disorder with 40 involved cases and vascular steal syndrome
with 26 involved cases were the most common complica-
tions.
The type of complications in our centers were different
from other centers, also, we had a high percentage (67.6%)
of AVF complications. Recombinant EPO is widely used to
treat anemia in hemodialysis patients, although different
side effects, including hypertension and thrombosis, have
been reported.13 On the basis of its prothrombotic effect,
EPO therapy has been suggested as a potential cause for
access dysfunction.13 However, our results did not confirm
this hypothesis. Although the low incidence of this compli-
cation among our patients biased the results, it would seem
to warrant further study to define its role in the pathogen-
esis of AVF thrombosis.
In our patient population, other well-established pro-
atherogenic factors, such as hypercholesterolemia, hyper-
tension, and diabetes, did not influence the risk of AVF
complications. Deposition of lipid droplets as well as the
presence of foam cells within the hyperplastic intima are
key features of the atherosclerotic lesions,13 but not of
venous stenosis of the AVF. Thus, it was not surprising that
we did not observe any association between elevated cho-
lesterol and AVF. The diabetic milieu is a powerful fibro-
genic stimulus, which induces a variety of micro- and
macrovascular damages. Although theoretically diabetes
can play a key role in AVF pathology, we did not demon-
strate any significant association between diabetes and AVF
dysfunction.
There are several justifiable reasons for protracted cath-
eter use in our center: patients awaiting living donor kidney
transplantation or maturation of a fistula. Catheters may
also be used as salvage for patients with no other potential
form of access. However, although these heroic uses are
increasing, they are still relatively uncommon. The reason
for the continued high prevalence of catheter use is unclear.
Possible explanations include decreased expertise with gen-
SALAHI, FAZELZADEH, MEHDIZADEH ET AL
eration of autologous fistulae within the surgical commu-
nity; misconceptions of unique complications (like high-
output congestive heart failure) associated with fistulae;
absence of profiling for vascular access types; inadequate
self-scrutiny of performance, and deliberate selection of a
catheter because of the ease of placement.14 It has been
suggested that early nephrology referral can facilitate au-
tologous fistula placement by providing time for fistula
maturation.15
REFERENCES
1. Denker BM, Chertow GM, Owen WF: In: Brenner & Rec-
tor’s kidney hemodialysis, 6th ed. Vol III. Philadelphia: Saunders;
2000, p 2377
2. Mehigam JT, Mc Alexander RA: Snuffbox arteriovenous
fistula for hemodialysis. Am J Surg 143:252, 1982
3. Feldman HI, Held PJ, Hutchinson JT, et al: Hemodialysis
vascular access morbidity in the United States. Kidney Int 43:1091,
1993
4. US Renal Data System: USRDS 1999 Annual Data Report.
Bethesda, MD: National Institutes of Health, National Institute of
Diabetes, Digestive and Kidney Diseases, 1999
5. Burkhart HM, Cikrit DF: Arteriovenous fistulae for hemodi-
alysis. Semin Vasc Surg 10:162, 1997
6. Mayers JD, Markell MS, Cohen LS, et al: Vascular access
surgery for maintenance hemodialysis: variables in hospital stay.
ASAIO J 38:113, 1992
7. Reyner HC, Besarab A, Brown WW, et al: Vascular access
results from the dialysis outcomes and practice patterns study.
Am J Kidney Dis 44(5 Suppl 3):22, 2004
8. Windus DW: Permanent vascular access: a nephrologist’s
view. Am J Kidney Dis 21:457, 1993
9. Turmel-Rodrigues L, Pengloan J, et al: Insufficient dialysis
shunts: improved long term patency rates with close hemodynamic
monitoring, repeated percutaneous balloon angioplasty, and stent
placement. Radiology 187:273, 1993
10. Glanz S, Gordon DH, Buff KMH, et al: The role of
percutaneous angioplasty in the management of chronic hemodi-
alysis fistulas. Ann Surg 206:777, 1987
11. Dapunt O, Feurstein M, Rendl KH, et al: Transluminal
angioplasty versus conventional operation in the management of
haemodialysis fistula thrombosis: results of a 5 year study. Br J Surg
74:1004, 1987
12. Schwab SJ, Raymond JR, Saeed M, et al: Prevention of
hemodialysis fistula thrombosis: early detection of venous stenosis.
Kidney Int 36:707, 1989
13. Grandaliano G, Teutonico A, et al: The role of hyperpara-
thyroidism, erythropoietin therapy, and CMV infection in the
failure of Arteriovenous fistula in hemodialysis. Kidney Int 64:715,
2003
14. Chertow GM: Grafts vs fistulas for hemodialysis patients:
equal access for all? JAMA 276:1343, 1996
15. Sommer BG, Sumrani NB, Hong JH: Current strategies for
vascular access in patients on hemodialysis. ASAIO J 38:743, 1992