The Michigan Vascular Access/MiDRE primer in Vascular Access

Chapter 10 - Options in hemodialysis access

There are three main options for hemodialysis – catheters, fistulas, and bridge grafts. Each has advantages and disadvantages. In the United States today there is a consensus that fistulas represent the optimal dialysis access in those patients who can have one, grafts are less desirable but acceptable, and that catheters should be limited to short-term use, or for those who cannot have a shunt (fistula or graft) due to limitations in cardiac reserve, poor peripheral arterial supply, or depleted venous assets.


A tunneled, cuffed catheter is frequently placed for dialysis access. These are popularly called "permacaths", although they are by no means usually permanent. Rather they should be distinguished from simple, non-tunneled, non-cuffed catheters popularly called "Quinton catheters" which are very temporary, and should not be used for more than 3 weeks. It is common to read "Quinton catheter" in the medical record when the patient actually has a tunneled, cuffed catheter, and vice versa.

A permacath with cuff
Up to %%% of patients initiating access in the United States find themselves with permacaths due either to the sudden onset of a previously un-appreciated condition, or because of procrastination and lack of preparation. It is recommended that patients be referred to a surgeon when their creatinine clearance (GFR) drops below 25 ml/min, but it is more common to see patients in the office with Gars in the 9 to 17 ml/min range. In this instance, one must place a usable shunt quickly, or the patient will need a catheter.

The advantage of a catheter is that it can be placed and used the same day. Because the hubs of the double lumen catheter are hooked directly to the dialysis machine tubing, there are no needles, no painful cannulation, and no cannulation problems. It is no wonder that many uninformed patients state that "the catheter is not bothering me", and delay getting a shunt, particularly seeing other patient’s in the unit with aneurysmal accesses, or accesses complicated with frequent clotting, or accesses complicated with prolonged bleeding.

The disadvantage of permacaths is that they break the skin, and render the body more vulnerable to infection. Up to 25 % of all dialysis patients will die of infection. Catheter related sepsis is a major risk and cost driver, but endocarditis and osteomyelitis are also devastating when they occur.

An underappreciated consequence of prolonged catheter use is the risk of central venous thrombosis and stenosis. One of my young patients nearly died when a huge thrombus behind a left internal jugular catheter was allowed to migrate into the lung as a pulmonary embolus. More often, the veins stenose, causing venous hypertension and arm swelling when that arm is used for a dialysis access (see below), or they become unusable for either a shunt or for a permacath, slowly whittling away at the options remaining for dialysis access (see chapter 28 on central stenosis).

A central venous catheter has occluded the right innominant vein

A patient and his femoral catheter

Venous hypertension and a swollen arm

Totally occluded veins from the left arm – now no longer usable

Ultimately, permacath usage doubles the risk of death in dialysis patients, and for this reason there is a renewed push to get the catheters out.

Fistulas A fistula is a vein that has been attached to an artery. The increased pressure and flow in the vein generally causes it to grow larger and tougher. When the vein is big enough, it can be used for dialysis. This process may take months. The failure of fistulas to develop to usability can be up to 50%. Maturation procedures such as dilating narrowings in the vein, bringing the vein up under the skin from deeper in the arm, or ligating diverting branches may be required. Fistulas are usually more irregular and more difficult to use than grafts. Fistulas sometimes grow to be unsightly (aneurysmal)

In order to have a fistula, you must have a vein that can be developed, a good enough artery to supply the blood flow necessary, and a surgeon who is able to put the two together. Patients with very bad hearts or bad circulation may not be able to tolerate the strain of a fistula.

The problems with fistulas are listed above. The advantages of a fistula are that they can last much longer than a graft, which may wear out in a few years, they "clog up" less often that grafts, and they are more resistant to infection. Fistulas are felt to confer a survival advantage compared to grafts in some patients, and definitely over catheters.

There are many types of fistulas, generally broken up into two types: "simple" and "complex" (see chapters 11 through 17)

A thirty year old wrist fistula which has never required intervention – a rarity
What is a graft?

A graft is a manufactured tube used to connect an adequate artery (one big and healthy enough to provide decent blood flow) and a healthy outflow vein capable of handling at least 1000cc of blood per minute. These artificial blood vessels are used to "bridge" the gap between artery and outflow when the patients other veins have been used up. Currently in the USA about 40% of patients dialyze using a bridge graft.

Grafts are more prone to clotting and infection than fistulas, and the average lifespan of a graft (three years) is less than a fistula, so in general fistulas are preferred where they are possible. However, a well-functioning graft may be preferable to a poorly functioning fistula. A well-functioning graft is much better than a catheter due to the decreased risk of infection.

Grafts have other important advantages: they can be placed right below the skin in a number of configurations nearly anywhere desired in the extremity, they are a uniform manufactured size and easier to use than many fistulas, they do not require maturation procedures or growth time and can be used much more quickly than fistulas. For that reason, graft placement can be an important part of the strategy for getting the catheters out.

Diagrams of an arm graft, upper arm loop and forearm loop graft
Grafts generally used in the US are made of expanded polytetrafluoroethelene (e-PTFE – the majority) in many variations of size and thickness, with tapered leading sections to limit flow, and trumpeted trailing sections to minimize turbulence where the graft is attached to the vein. There are carbon lined grafts and heparin bonded e-PTFE grafts. There are grafts with a stent built into the trailing end. There are polyurethane grafts that can be used the next day after placement, "trilaminant" e-PTFE grafts that can be used within three days and biological grafts made from bovine mesenteric vein or carotid artery, all cured and sterilized to prevent rejection by the body.

My experiences with biological grafts include a number of cases with the Hanson-Jaffe Procol Bovine Mesenteric Vein graft, and 100 cases with the Artegraft Bovine Carotid Artery graft. Expectations that these biological grafts would be more like a fistula than a graft and lead to better overall outcomes were not born out in my experience, and I am no longer using these conduits.

Currently, I am placing grafts when patient have no fistula options, or when the chances of maturing a fistula before the patient has to go on dialysis are slim. A graft in the forearm is sometimes used as a temporary option to help develop the veins in the upper arm for fistulas later. The graft can be used in short order (getting the catheter out before it can cause damage to be veins), and down the line a transition from graft to fistula may be possible (see Chapter 18).

A forearm loop graft which has produced a sizable outflow vein

In conclusion, grafts will continue to have an important role in providing access for dialysis as long as patients have inadequate veins, or veins that have been destroyed, or come to dialysis too late to establish a fistula in a timely fashion. The realization that catheters begin to create irreversible changes much earlier than previously believed has pushed for the removal of catheters within 90 to 150 days (the "ninety-day cath-away" and "150 day cath-away" programs). This will only be a reality when surgeons are able to create and mature a fistula within a short time, and know when a graft is the better option.

In practice

Practically speaking, provision of dialysis access depends on two critical steps: (1) deciding what kind of access is possible (judgment), and (2) the technical ability to perform the surgery (skill). Both come from experience, but of the two, judgment is probably the most important.

In 2010, 326 patients new to our practice were seen: 58 pre-dialysis patients (18%), 124 catheter dependent, new-to-dialysis patients needing first-time access (38%), 79 patients with previous failed access needing new access (24%), and 65 patients with problematic accesses performed elsewhere needing intervention (20%). In addition, over 448 established patient visits were scheduled for problems with existing access. Over 1200 open and endovascular procedures were done in 2010.

Overall, 285 new accesses were placed, with 242 fistulas and 43 grafts, for an overall fistula rate of 85%. Just creating fistulas is not enough however, since widely reported national rates of fistula maturation are in the 50-65% range. In order to assess the success of a vascular access program one must know raw creation rate (the 85% noted above), the maturation rate, and the time to usability.

Results on fistulas

In 2011 I generated some statistics from examining my own practice and found that maturation rates varied from 78% (wrist fistulas) to 96% (transposed basilic fistulas), and that most fistulas could be cleared for use from an average of 62 days (one-stage brachiocephalic fistulas) to 113 days (two-stage transposed basilic fistulas).

The value of these data are threefold: first, that the 50-60% maturation rate of fistulas reported nationally can be bettered in focused surgical practices focusing on dialysis accesses, meaning that a 50/50 success rate should no longer be considered acceptable; secondly that a 80-95% success rate rather than 50% means that multiple access attempts should be needed less often, getting the catheters out faster; and finally, the fact that normal fistulas can be matured in less than three months means that knowledgeable aggressive follow-up does pay off, and that procrastination does not.
Percent matured Time to release
Wrist fistulas 78% 72 days
Forearm fistula 83% 75 days
Antecubital fistulas 85% varies
Brachiocephalic AVF 85% 62 days
Superficialized BCAVF 95% 102 days
Transposed basilic AVF—one stage 96% 50 days
Transposed basilic AVF—otwo Stage 98% 113 days

Knowing that two stage basilic transpositions take an average of four months to reach usability has pushed me to do more one-stage procedures, or to shorten the interval between stages to save time on the catheter and reduce central venous damage.

Knowing that wrist fistulas have the highest failure rate, I have been more critical about who is a real candidate for a distal fistula and who is not. There is no point to "trying" a fistula which is not destined to work (see image below).

Knowing that even in the best of hands the recommendation is for a fistula 85% of the time, and only 85% of fistulas mature, it follows that 72% of patients will need something else, either a better fistula, or a graft. If a catheter is present, or time is pressing, I am more critical about choosing a low yield fistula rather than a higher yield graft. The critical issue is really getting the catheters out.

A poorly performing fistula due to inadequate arterial supply – something to avoid

Ultimately, options are determined by a patient’s health status and anatomy, plus the surgeon’s experience. Recommendations are driven by these factors, plus patient preference and the pressure of time when a catheter is present, or when dialysis is imminent.

Michigan Dialysis Research and Education (MiDRE)
Saint Mary's Mercy Hospital