CHAPTER 4

Abstract

I. Planning for Vascular Access Recommendations Each center should establish a dedicated team for vascular access. (Grade D, opinion) Preserve arm veins suitable for placement of vascular access. Preservation should begin in patients with progressive kidney disease and an estimated GFR of less than 30 ml/min. (Grade D, opinion) The preferred type of vascular access is a radio-cephalic native vessel arteriovenous fistula. (Grade C) Background Arteriovenous (AV) access–related complications result in considerable morbidity. With a dedicated access team, including a nephrologist, an access surgeon, an interventional radiologist, and a dialysis nurse, a center can develop and maintain skills that should lead to better patient care. Arm veins, particularly the cephalic veins of the nondominant arm, should not be used for venipuncture or intravenous catheters. In patients with advanced chronic kidney disease, the dorsum of the hand should be used for intravenous line. When venipuncture of the arm veins is necessary, sites should be rotated. Patients should wear a Medic Alert bracelet to inform hospital staff to avoid intravenous cannulation of essential veins. The preferred type of access is a native AV fistula, followed by grafts and then central venous catheters (1–5). It has been shown that the relative risk of bacteremia is greater with central venous catheters than with AV fistulae (6). Compared with AV grafts, AV fistulae have been shown to be associated with better long-term survival, require less intervention to maintain patency, have lower infection rates, and lead to less health care expenditure (1–5,7). Data from the Canadian Organ Replacement Registry show equivalent patient survival on hemodialysis and peritoneal dialysis (8). This information should be considered before using central venous catheters for long-term hemodialysis as opposed to peritoneal dialysis, when both are technically feasible. The preferred sites for placing the AV fistula are (in order of preference) the wrist (radiocephalic) and the elbow (brachiocephalic) (3,9,10). If it is not possible to establish either of these types of fistula, access may be established using either a transposed brachial-basilic vein fistula (3) or an AV graft of synthetic material (e.g., polytetrafluoroethylene [PTFE]). The preferred site and type of graft is a forearm curved looped radiocephalic graft. This is followed by an upper arm straight graft (11). The least preferred sites and type of grafts are forearm straight radial cephalic and looped thigh grafts. However, the location for the graft placement is determined by each patient’s unique anatomical restrictions, previous access history, and the surgeon’s skill. During the planning phase for a new vascular access, venography may be beneficial and may be considered in patients with the following: edema in the extremity in which an access site is planned, collateral vein development or accessory vein enlargement in any planned access site, differential extremity size, if that extremity is contemplated as an access site, current or previous subclavian catheter placement of any type in venous drainage of planned access, current or previous transvenous pacemaker in venous drainage of planned access, previous arm, neck, or chest trauma or surgery in venous drainage of planned access, or multiple previous accesses in an extremity planned as an access site. For patients not yet on dialysis, the benefit of the information gained with venography must be weighed against the risk associated with exposure to radio-contrast media. II. Access Timing, Placement, and Maturation Recommendations Establish AV fistulae when the patient has an estimated GFR of 15 to 20 ml/min and progressive kidney disease. (Grade D, opinion) Background It is important to create an AV fistula at least 3 to 4 mo before its anticipated use. More time may be necessary depending upon the site’s referral and surgical wait times. Grafts can be used in patients who are not candidates for a primary AV fistula. Place dialysis PTFE AV grafts at least 3 to 6 wk before an anticipated need for hemodialysis. According to the available evidence, PTFE tubes are preferred over bovine grafts (12). There is no evidence available yet on the efficacy of newer synthetic materials. When using these materials, follow manufacturers’ recommendations. Cuffed and noncuffed hemodialysis catheters can be inserted immediately before their use because they do not require maturation time. Cuffed, tunneled, central venous catheters can be a valuable alternative to grafts, although there are concerns about infection, thrombosis, and dialysis adequacy. Adjust the catheter tip to the level of caval atrial junction or beyond. Subclavian access should be used only when jugular options are not available. Catheter position should be confirmed using radiography, and the catheter tip should be readjusted as necessary to ensure proper position. The use of real-time, ultrasound-guided insertion may be an advantage in reducing insertion-related complications, particularly in patients who have had previous catheter insertions (13,14). Do not place jugular or subclavian hemodialysis catheters on the same side as a maturing AV access. For patients with chronic kidney disease who need acute hemodialysis vascular access, use a noncuffed or a cuffed percutaneously inserted catheter. These catheters are suitable for immediate use and should not be inserted before needed (15). Femoral catheters should be at least 19 cm long to minimize recirculation. Noncuffed femoral catheters should be sutured in place and can be left in as long as there are no complications. AV fistulae need time to mature before cannulation (at least one month, preferably 3 mo). Recent data from the Dialysis Outcomes and Practice Patterns Study (DOPPS) demonstrates a large intercountry variability in the timing of AV fistula cannulation. The majority of fistulae in Europe are cannulated early—within 8 wk of creation. Early cannulation does not appear to be associated with subsequent fistula failure and may decrease exposure time to central venous catheters (16). As AV fistula maturation depends on artery and vein size and integrity as well as cardiac output, clinical judgment should be used in determining time to first use. The following procedures have been used in an attempt to enhance maturation of AV fistulae: fistula hand-arm exercise (e.g., squeezing a rubber ball with or without a lightly applied tourniquet), selective obliteration of major venous side branches, rest, until swelling is resolved (for a new native AV fistula with induration and edema). A new PTFE dialysis AV graft should not be cannulated until swelling has gone down enough to allow palpation of the course of the graft—ideally 3 to 6 wk after placement. Ideally, no attempt should be made to cannulate the graft for at least 14 d after placement. Use a venogram or other noncontrast study to evaluate central veins in patients with swelling that does not respond to arm elevation, or that persists >2 wk after dialysis AV access placement. III. Monitoring and Maintenance of Vascular Access Recommendations Measure access flow bimonthly in AV fistulae (Grade D) and venous pressure or access flow monthly in AV grafts. (Grade D) Perform angiography if fistula flow decreases to <500 ml/min or drops >20% from baseline (Grade D); if AV graft flow decreases to <650 ml/min or drops >20% from baseline. (Grade D) Background Monitoring AV fistulae and grafts for hemodynamically significant stenosis, combined with corrective treatment, improves patency and decreases the incidence of thrombosis (17–24). A quality assurance program should collect and maintain data on each patient from the monitoring tests, clinical assessment, and dialysis adequacy measurements, and make this information available to all staff. The data should also be tabulated and tracked within each dialysis center and benchmarked against regional or national standards. Although recirculation studies have been shown to be useful for detecting AV fistulae stenosis, the recirculation only occurs when the total access flow is lower than the blood flow in the dialysate circuit. Therefore, the preferred method for monitoring AV fistulae is direct on-dialysis flow measures (18). When clinicians do not have access to on-dialysis flow measures, they can monitor AV fistulae using regular recirculation studies (25). When using flow measures, clinicians should be aware that AV fistulae are capable of sustaining a lower blood flow than an AV graft without clotting, so a flow measurement <650 ml/min in an AV fistulae is less likely to indicate a reversible stenosis or subsequent clotting. However, relative changes in flow measurement are still a cause for concern. After a successful angioplasty, AV fistulae should be monitored monthly and investigated if a flow <500 ml/min or a drop of access flow >20% of baseline occurs (18–20). When using access recirculation measures, clinicians should be aware that any access recirculation is abnormal and should be investigated. Recirculation >5% using non–urea-based methods and recirculation >15% measured using urea-based method is significant and should lead to angiography. Methods for monitoring AV grafts include: intra-access flow including monitoring for changes in flow (26–33), static venous pressures (25), dynamic venous pressures (34), slow-flow venous pressure (35). Blood access flows through AV grafts can be measured by indicator dilution or conductivity tracer techniques, using the Krivitski reversed line technique (28). In a prospective study of 170 chronic hemodialysis patients, May et al. demonstrated that access blood flow measurements were superior in the prediction of access thrombosis compared with static pressure monitoring or urea recirculation measurement (30). However, in a blinded, randomized, controlled trial of AV graft monitoring and angioplasty, monthly blood flow measurement did not improve graft thrombosis rate over and above the standard surveillance (dynamic venous pressure and physical examination) (36). When using pressure measurements to monitor access, clinicians should be aware that static pressure measurements are more accurate than dynamic pressure measurements (25). Methods to measure dynamic, static, and slow-flow venous pressures are provided in the Table 2 Other studies or information that may be useful in detecting AV graft stenosis include: measurement of access recirculation using urea concentrations; measurement of recirculation using dilution techniques (non–urea-based); unexplained decreases in the measured amount of hemodialysis delivered (urea reduction ratio, Kt/V); physical findings of persistent swelling of the arm with the graft, prolonged bleeding after needle withdrawal, or altered characteristics of pulse or thrill in a graft; elevated negative arterial prepump pressures that prevent increasing to acceptable blood flow; venography/Doppler ultrasound. Any finding of access dysfunction, whether based on the presence of access recirculation, low or deteriorating access blood flow rates, positive pressure tests, or any other test should be investigated using angiography to determine the appropriate intervention (e.g., angioplasty, surgery). In the process of investigating the dysfunction and taking corrective measures, it is vital that the clinician take interim measures to protect the patient. When the dialysis circuit blood flow exceeds the access flow, access recirculation will occur, which leads to inadequate dialysis. To optimize dialysis treatment, the dialyzer blood flow should be reduced to a level at or just below the patient’s measured access blood flow rate. The clinician should make the appropriate corrections by time and dialyzer surface area to ensure that the patient receives the desired and prescribed Kt/V (urea). IV. Infection Prevention in the Vascular Access Recommendations Instruct all staff and patients on infection control measures. (Grade D, opinion) Change catheter exit site dressings at each hemodialysis treatment (Grade D, opinion). Use dry gauze dressings and povidone iodine (Grade C), mupirocin (Grade C), or polysporin triple ointment (Grade A) at the catheter exit site. Background Proper infection control procedures can significantly reduce the risk of infection. Catheter care and accessing the patient’s circulation should be sterile procedures. During catheter connect and disconnect procedures, nurses and patients should wear a surgical mask or face shield. Nurses should also wear gloves during all connect and disconnect procedures, although the evidence for sterile versus nonsterile gloves is inconclusive. Use a clean technique for needle cannulation for all cannulation procedures. Ensure that only trained dialysis staff or caregivers change hemodialysis catheter dressings and manipulate catheters that access the patient’s bloodstream and minimize contamination. A randomized control trial of dry gauze dressing with povidone iodine ointment at the catheter exit site, along with sterile dressing technique, resulted in a significant reduction in Staphylococcal aureus exit site infections, bacteremia, and catheter tip colonization (37). The beneficial effect was most evident in S. aureus carriers. Similar results have recently been reported using mupirocin and polysporin triple ointment (38). In the recent study by Lok et al., polysporin triple ointment was associated with a survival benefit (39). Routine monitoring for staphylococcal nasal carrier status and its management remains controversial. Although some studies have shown reduction in S. aureus bacteremia in hemodialysis patients with nasal mupirocin ointment, development of antimicrobial resistance remains an important concern. V. Managing Vascular Access Complications Recommendations Use percutaneous angioplasty to treat all hemodynamically significant stenoses in patients with AV fistulae and AV grafts (Grade D); if percutaneous angioplasty is not possible, use surgical revision. In the case of AV fistulae aneurysm formation, surgically intervene if the skin overlying the fistula is compromised, the aneurysm is expanding, or available puncture sites are limited. (Grade D) In the case of AV grafts, surgically intervene in the presence of graft degeneration and pseudoaneurysm formation. (Grade D) Correct thrombosis of an AV graft with pharmacomechanical or mechanical thrombolysis or surgical thrombectomy. (Grade D) Background Angioplasty is the preferred treatment for both fistulae and graft stenosis (21,22,34,40,41). In native vessel AV fistulae, the most common site of stenosis/thrombosis is near AV anastomosis, distal to the insertion of an arterial needle (40,42). Stenosis, as well as the clinical parameters used to detect it, should return to within acceptable limits after the intervention. Centers should monitor stenosis treatment outcomes on the basis of patency. It is this committee’s opinion that reasonable patency goals (for the center as a whole) for angioplasty and surgical revision in the absence of thrombosis are: Angioplasty: 50% unassisted patency at 6 mo; for all patients, no more than 30% residual stenosis postprocedure and resolution of physical indicator(s) of stenosis. Surgical revision: 50% unassisted patency at 1 yr. If angioplasty is required >2 times within 3 mo and the patient is a good surgical candidate, referral for surgical revision may be consdiered. Stents are useful in selected instances (e.g., central venous stenosis, limited residual access sites, surgically inaccessible lesions, contraindication to surgery) when angioplasty fails. The choice of technique to correct thrombosis should be based on the center’s expertise. Treatment should be performed as rapidly as possible (within 24 h) after detection of thrombosis to minimize the need for temporary access. The access should be evaluated by fistulogram for residual stenosis postprocedure. Residual stenosis should be corrected by angioplasty or surgical correction. Outflow venous stenoses are present in >85% of instances of thrombosis of AV grafts. The need for percutaneous transluminal angioplasty or surgical revision is expected in most instances. Monitoring tests used to screen for venous obstruction should return to normal after the intervention. Centers should monitor outcome results on the basis of patency. It is this committee’s opinion that minimum reasonable goals (for the center as a whole) for percutaneous thrombolysis and surgical revision thrombectomy should be: Percutaneous thrombolysis with angioplasty: 40% unassisted patency and functionality at 3 mo. Surgical thrombectomy and revision: 50% unassisted patency and functionality at 6 mo and 40% unassisted patency and functionality at 1 yr. Immediate patency (patency to next hemodialysis session): 85% for both techniques. Prophylaxis of access thrombosis has not been extensively studied. A recent, randomized, double-blind, placebo-controlled trial of coumadin versus placebo failed to demonstrate a difference in thrombotic events in newly placed AV grafts. Major bleeding was also more common in patients assigned to coumadin therapy (43). There is also some question whether the widespread use of acetylsalicylic acid may be counterproductive (44). Although a small, randomized trial has provided some enthusiasm for the use of fish oil in the prevention of AV graft thrombosis (45), additional trials are required before widespread use of this therapy could be recommended. Treat hand ischemia from arterial steal with a distal revascularization internal ligation procedure. If this fails or is not feasible, consider ligation of the AV fistula or graft. (Grade D) Treat symptomatic central vein stenosis with percutaneous transluminal angioplasty. Place a stent only after failed angioplasty. (Grade D) Treat dysfunctional tunneled hemodialysis catheters with instillation/infusion of tPA using a protocol. (Grade D, opinion) Background Significant hand ischemia occurs in 2 to 8% of patients with AV access. Risk factors include female sex, age >60 yr, diabetes, and use of brachial artery as a donor vessel. A 10-yr retrospective review of this complication in Athens, Greece, revealed 28 of 569 patients with proximal AV access developed this complication (46). The Distal Revascularization Interval Ligation procedure, which includes an arterial ligation placed just distal to the AV graft or anastomosis and short bypass from a point 4 to 5 cm proximal to the inflow of the access to a point just distal to the ligation, was performed in 23 of these patients. Immediate relief of symptoms occurred in all and 1-yr patency was 69% (46). Central vein stenosis can result in significant arm swelling when an AV access is created on the ipsilateral side. When a patient has central vein stenosis and significant arm swelling, percutaneous angioplasty should be performed. Angioplasty can be repeated in case of recurrence. A stent should be placed after more than one recurrence or a failed angioplasty (47). A protocolized approach is recommended for management of a dysfunctional hemodialysis catheter. Catheter dysfunction is defined as failure, based on catheter-related thrombotic or mechanical factors, to attain and maintain an extracorporeal blood flow sufficient to perform the prescribed hemodialysis without significantly lengthening or altering the hemodialysis treatment. A common cause for this dysfunction is the development of a fibrin sheath around the catheter, which can develop shortly after insertion. The fibrin sheath acts as a nidus for thrombus and biofilm formation. With the unavailability of urokinase, tissue plasminogen activator (tPA) has been used to restore catheter patency reduced by thrombosis. Instillation of 2 mg tPA in each lumen for a median of 24 h resulted in patency in >80% of cases (48). Given the available information, treatment of a dysfunctional permanent catheter using a protocol for tPA instillation is recommended (Table 3). If adequate flow is not achieved, consideration can be given to the use of low-dose systemic tPA. Although there is no specific trial, this recommendation is based on systemic urokinase use (49). The reader should keep in mind that valid evidence to support the manner of tPA instillation (dwell versus advancing versus infusion protocols) and the duration of the therapy is lacking. If adequate flow is still not achieved, perform radiographic studies followed by intervention. In patients with contraindications to systemic administration of tPA, fibrin sheath stripping (50) and catheter exchange over a guidewire are alternative options. Treat extensive infection of a dialysis AV graft with parenteral antibiotics and total graft resection. (Grade D) Treat infections of primary AV fistulae as subacute bacterial endocarditis with 6 wk of antibiotic therapy. (Grade D) Treat central venous catheter-related bacteremia with systemic antibiotics and catheter exchange over a wire. (Grade D) Treat catheter tunnel infections without bacteremia with parenteral antibiotics and appropriate local measures. Catheter removal is indicated if the infection fails to respond to 2 wk of therapy. (Grade D) Background Use of central venous catheters is associated with a significantly higher risk of bacteremia compared with AV fistulae. With AV grafts, the infection risk is moderate (1,6,51). It may be possible to eradicate a local graft infection with a combination of incision and local resection of the infected portion of the graft and systemic antibiotics (52). However, extensive infection of a graft requires total resection of the graft along with parenteral antibiotics. Tunneled cuffed catheter infection is a serious problem. Appropriate treatment depends on the nature of the infection (53–55). In patients with cuffed or noncuffed central venous catheters and suspected bacteremias, AV fistulae, or AV graft infections, start treatment with 1 to 2 g cefazolin depending on patient weight, and 1.5 mg/kg gentamycin postdialysis after blood cultures are drawn. In patients with known cephalosporin allergy, or in centers with a predominance of coagulase negative staphylococcal catheter related infections, use 15 mg/kg vancomycin instead of cefazolin. Once blood culture results are available in stable asymptomatic patients without exit site or catheter tunnel tract involvement, the catheter should be changed over a wire and antibiotic treatment continued for 2 to 4 wk as clinically indicated. In all cases, definitive therapy should be based on the organism(s) isolated. For patients with central venous catheters, the catheter should be removed, rather than exchanged, in all instances if the patient is clinically unstable or if the patient remains symptomatic for >36 h (53–55). In patients with difficult access, clinicians can attempt antibiotic treatment without changing the catheter. However, the success of such catheter salvage is low (54,55). A new, permanent access should not be placed until blood cultures, performed after cessation of antibiotic treatment, have been negative for at least 48 h. Catheter exit site infections are characterized by redness, crusting, and exudate at the exit site in the absence of systemic symptoms and negative blood cultures. Treatment includes proper local exit site care and oral or parenteral antibiotics based on culture and sensitivity. The catheter typically does not need to be removed. If there is tunnel drainage, treat with parenteral antibiotics (antistaphylococcal or antistreptococcal therapy pending culture report) in addition to following appropriate local measures. Definitive therapy should be based on culture results. Do not remove the catheter unless the infection fails to respond to therapy or the patient is clinically unstable. If the infection fails to respond after 2 wk of therapy, remove the catheter and replace it using a different tunnel and exit site. For all access-related infections, empiric therapy should be regularly evaluated in conjunction with specialists in infectious diseases or microbiology so that therapy reflects the changing microbiology and sensitivities that are unique to the local environment. VI. Quality of Care Standards Background Primary AV fistulae should be constructed in all suitable new patients who elect to receive hemodialysis as their initial form of renal replacement therapy. After failure of every dialysis AV access, all patients should be re-evaluated for possible construction of a primary AV fistula. Ultimately, >60% of prevalent patients should have a native AV fistula (1). Each center should establish a database to track the types of accesses created and the complication rates. Centers should work to achieve the following target rates: The rate of graft thrombosis should not exceed 0.5 thrombotic episodes per patient year at risk (34). After adjusting for initial failures (e.g., failures within the first 2 mo of fistula use), the rate of thrombosis of native AV fistulae should be <0.25 episodes per patient year at risk. Dialysis centers should examine their thrombosis rates and the underlying causes as part of an ongoing Quality Assurance/Continuous Quality Improvement program. The rate of infection should not exceed 0.01 episodes per patient year at risk for primary AV fistula and 0.1 episodes per patient year at risk for AV grafts (6). For tunneled cuffed catheters, the recommended target rate of systemic infection is <0.5 episodes per patient year at risk (6). The primary access failure rates of dialysis AV grafts in the following locations and configurations should not be >15% in forearm straight grafts, 10% in forearm loop grafts, and 5% in upper arm grafts (3,11,56). The cumulative patency rate of all dialysis AV grafts should be at least 70% at 1 yr, 60% at 2 yr, and 50% at 3 yr (40,41). Recommendations for Research To improve the clinicians’ ability to monitor and intervene successfully, future research should include a randomized prospective trial on intervention based on access flow measurements in both AV fistulae and PTFE grafts. Randomized trials should be performed to assess the suitability of clinical and vascular studies before AV access creation to improve AV access maturation. Appropriate trials for prevention of venous thrombosis in AV grafts should be performed. Appropriate studies to determine characteristics that would influence/predict the successful creation and maturation of fistulas are required. Table 2: Methods to measure dynamic, static, and slow-flow venous pressuresTable 3: Algorithm for malfunctioning central venous catheter