Surveillance and Management of Failing Hemodialysis Access

Indications

All patients who have hemodialysis fistulas or grafts should have periodic surveillance of the dialysis access flows. The KDOQI guidelines recommend an organized monitoring/surveillance approach with regular assessment of clinical parameters of the dialysis access and hemodialysis adequacy. Data from the clinical assessment and hemodialysis measurements should be collected and tracked as part of a quality assurance program. In the KDOQI guidelines, monitoring refers to examination and evaluation of the vascular access by means of physical examination to detect physical signs that suggest the presence of dysfunction. Monitoring includes physical examination of the vascular access, review of routine laboratory studies, dialysis adequacy, difficulty in cannulation or achieving hemostasis after needle removal, and documented recirculation.

Surveillance refers to periodic evaluation of the vascular access by using tests that may involve special instrumentation and for which an abnormal test result suggests the presence of dysfunction. Access flow measurement, duplex Doppler ultrasound, and direct or derived static pressure are frequently used surveillance tools.

Diagnostic testing refers to specialized testing prompted by some abnormality or other medical indication and undertaken to diagnose the cause of the vascular access dysfunction.

Physical examination should be performed on all fistulas and grafts at least monthly to detect dysfunction; this is mandated by Medicare. Physical examination is useful, reliable, easily performed, and inexpensive. A number of techniques may be used for surveillance of grafts. Dynamic venous pressure is usually considered a monitoring strategy rather than surveillance because dialysis machines measure the dynamic venous pressure during treatment. However, the utility of dynamic venous pressures at flows of 150 to 200 mL/min in detecting stenosis or predicting access thrombosis is limited. Because of this, dynamic venous pressure measurements are not recommended for surveillance. The most common techniques that are truly surveillance tests include access blood flow (ABF), direct or derived static venous pressure (SVP) measurements, and duplex Doppler ultrasound.

Abnormal values include a flow rate less than 600 mL/min in grafts and less than 400 to 500 mL/min in fistulas, a static venous pressure ratio greater than 0.5 in grafts or in fistulas, or an arterial segment static pressure ratio greater than 0.75 in grafts. The values of ABF or SVP can vary during dialysis session. Cannulation techniques and changes in hemodynamics during a dialysis session can alter these values, so one should not necessarily respond to a single abnormal value. Because flow and change in flow are inaccurate predictors of thrombosis, several different measurements should be obtained over time. If a trend is noted, the patient should be referred for diagnosis and treatment. This seems to be particularly important in new grafts with low flow and a decrease in flow because these findings seem to be predictive of graft failure.

Contraindications

There are no contraindications to dialysis access surveillance. Angiographic diagnosis and treatment of the dialysis access may be contraindicated in some patients and would include those associated with contrast media, primarily severe allergic reactions. Other contraindications would be systemic infections, particularly in the case of a thrombosed access where the thrombus may be infected.

Early failure is usually due to a technical error in creation of the fistula or graft. Rarely, no underlying anatomic lesion can be identified. These patients may have had excessive postdialysis compression, hypotension, hypovolemia, graft compression due to sleeping position, or a hypercoagulable state, any of which may lead to thrombosis.

Monitoring Examination

The KDOQI Venous Access Working Group strongly encourages use of physical examination for monitoring dialysis fistulas and grafts. There is a concern that the basic skills of inspection, palpation, and auscultation have been largely abandoned in favor of technology, and the Working Group believes these skills should be taught to all individuals who perform hemodialysis procedures. Physical findings of persistent swelling of the arm, presence of collateral veins, prolonged bleeding after needle withdrawal, or altered characteristics of pulse or thrill in the access suggest the possibility of access dysfunction. Simple inspection can reveal the presence of aneurysms. Flow through the access can be evaluated by assessing the thrill (vibration) at various points along the access. When there is a stenosis present, a “water-hammer” pulse is felt below the stenosis, and above the stenosis the pulse goes away abruptly. If accessory veins have developed, they may frequently be visible, and the pulse in the main outflow vein decreases above the level of the side branch. Auscultation of the graft, particularly at the region of the venous anastomosis, can be informative. When a stenosis is present, there is often a high-pitched, harsh, or discontinuous bruit at the affected site. Because the velocity of flow increases in an area of stenosis, there will be a localized bruit or localized increase in the pitch of the bruit. Marked arm swelling usually indicates the presence of a central venous stenosis. Multiple dilated veins and collaterals are usually apparent in the patient’s upper arm and chest, and swelling of the breast may also occur with central venous stenosis. If an abnormality is found on physical examination, about 90% of patients will be found to have an angiographically significant abnormality.

Surveillance Technology

Surveillance can be performed in a number of ways according to the KDOQI guidelines ( Box 80.1 ). Flow decreases by less than 20% until the stenosis is 40% to 50%, and then the decrease in flow is rapid as the degree of stenosis increases to 80%. Thus the rationale for surveillance is based on the hypothesis that progressive stenosis is accurately detected by reduced flows and increased venous pressure. Decreasing flows can be measured by access flow measurements. Access flow is the most common surveillance method and can be either direct or indirect. There are multiple methods for indirect measurement of access flows that have been predictive of access dysfunction. Most use some type of dilution technique. These include ultrasound dilution (Transonic, Ithaca, NY), timed ultrafiltration, ionic dialysance, differential conductivity, and glucose infusion. Access flow is measured by inducing forced recirculation, which is created by reversing the arterial and venous blood lines. The ultrasound dilution technique is the most well-validated method for indirectly measuring blood flow. A saline bolus is administered and the amount of dilution is measured by a downstream ultrasound sensor. This technique is advantageous in that it can be performed during dialysis to reduce time spent on the measurement. Access flow can also be monitored through changes in temperature, sodium, hemoglobin, and hematocrit. There are dialysis machines with embedded blood temperature monitors and monitors for hemoglobin or hematocrit (“crit lines”). Thermal dilution uses temperature changes measured by a sensor after the infusion of cold saline. In-line dialysance measures flow by injecting a small bolus of hypertonic saline in the venous line and measuring the change in conductivity in the arterial line. Other methods include infusion of glucose into the access and the glucose level downstream is compared with a baseline sample. This technique has poor reproducibility and requires multiple blood draws so it is not recommended. A change in ABF of >20% provided the best combination of sensitivity (86%) and specificity (90%) for graft thrombosis.

Access flow is useful for identifying inflow stenosis. However, variations in access flow during dialysis can result from changes in cardiac output, mean arterial pressure, and blood volume. Access flow can increase by up to 11% or decrease by up to 30% from initial values by the end of dialysis, potentially impairing the ability of flow to predict impending vascular access failure. A single abnormal value is thus less valuable than demonstrating a trend over time. When access flow is measured repeatedly, trends of decreasing flow add predictive power for detection of access stenosis or thrombosis.

Access pressures have also been used to evaluate dialysis access function. The pressure required to infuse blood back into the access is recorded as a venous pressure. A mean venous access pressure ratio can be calculated by simultaneous measurement of the static venous access pressure divided by the mean arterial pressure measurement. When a stenosis develops, the pressure increases and the flow decreases. One of the components of the venous pressure is the static pressure; when static pressure increases to greater than 50% of the mean arterial pressure, the graft flow commonly has decreased into the thrombosis-prone range of 600 to 800 mL/min, and the presence of a stenosis is likely. SVP is a better tool for detecting outflow lesions. Although the SVP measurement may work relatively well for a graft, in a fistula there may be collateral veins decompressing the fistula, making SVP measurements less valuable as a surveillance tool. It has been shown that SVP measurement has a 92% positive predictive value for significant stenosis in AVG. There are situations in which the SVP is not accurate, including arterial stenoses or stenosis between the area used for the arterial needle and the area used for the venous needle. In both instances, the SVP may be normal or even decrease despite increasing stenosis and decreased flows. Although measuring SVP is straightforward, it is tedious, time consuming and not “user friendly,” so it is not considered an ideal method for surveillance.

Recirculation—the return of dialyzed blood to the dialyzer—had been used as a measure of dialysis access function. It can be a sign of low ABF and a marker for stenosis. However, it is a very late manifestation of stenosis and a poor predictor of imminent thrombosis and is no longer recommended by the KDOQI guidelines. This technique may be used in fistulas, although it tends to be a late finding and should have a minor role.

Duplex Doppler ultrasound imaging is considered a direct method for evaluating access flows. It can be used to detect stenoses in both grafts and fistulas. The method is operator dependent and subject to error caused by variation in cross-sectional area and the angle of insonation. Duplex Doppler ultrasonography flow measurements do not have good sensitivity and specificity. Duplex ultrasound is much more accurate in performing anatomic assessment and direct evidence for the presence, location, and severity of access stenosis. It is particularly useful to determine reasons for maturation failure of fistulas. Because many fistulas cannot be studied using other surveillance techniques, routine duplex Doppler surveillance of primary fistulas should be considered to identify flow-limiting stenoses that may compromise the long-term patency and use of the fistula. In experienced hands it has an excellent correlation with fistulograms and facilitates the work by the interventional radiology team.

Magnetic resonance angiography is also considered a direct method for access evaluation. It can also be used to evaluate the dialysis access noninvasively and allows anatomic assessment. Magnetic resonance angiography is effective in characterizing stenosis and obtaining flow velocities. But its high cost and the potential for gadolinium-induced nephrogenic systemic fibrosis limit its use.