GUIDELINES [ back to index ]
10. Routine management of A/V fistula and graft
*1 Checks before each cannulation
• Inspection of the access site for signs of infection (redness, discharge, oedema), aneurysms, haematoma and colour (red, blue, white) of the hand is important for the early diagnosis of impending complications1. Swelling due to venous hypertension and the potential progression of acral necrosis should be monitored prior to haemodialysis treatment. Venous dilatations and aneurysms in grafts without stenosis (low pressure aneurysms) can be cannulated using the lateral side provided that the skin is completely healthy.
• Palpation along the access site from its arterial anastomosis up to the outflow veins may give an idea of the function of the access. In fistulas normally a strong thrill and weak pulsation is felt. In grafts the presence of a good thrill is associated with flows of > 450 ml/min 2. The intravascular pressure should be equal at any segment. Differences in intravascular pressure and an increase in pulsation indicates an outflow stenosis. In fistulas a stenosis can be felt along the access tract by palpation. Palpation of peripheral pulses (radial and ulnar at the wrist) is indicated in patients with hand perfusion disturbances (white/blue hand). However, in one third of proximal procedures with the brachial artery as in flow artery there is no distal pulse as a consequence of a mild non limb-threatening steal 3. Digital pressure or wrist-to-brachial pressure index measurements could be used instead. An index > 0.5 or a digital pressure > 50 mm Hg rules out a limb threatening steal 3 4.
• Auscultation normally reveals a typical thrill with a long diastolic component that is shortened in case of haemodynamically relevant stenosis. A stenosis causes a high-pitch bruit. In thrombosis, there will be no bruit at all.
• In patients with grafts, palpation may be difficult due to the deep subcutaneous position of the graft. Auscultation is more reliable and should concentrate on the venous outflow tract for detection of graft-to-vein anastomotic stenosis.
*2 Check during and after each cannulation
• Difficulties in access cannulation or prolonged bleeding should be documented in the patients’ medical records and addressed by the medical team 5.
*3 Ease of puncture
• After overcoming the initial resistance of the access wall, the fistula needle should be easily advanced into the access. Further resistance may indicate that the needle is not within the access lumen or that it may have entered the back wall. Resistance can also indicate the presence of pseudo-intimal hyperplasia or thrombus.
*4 Aspiration of clots
• Aspiration of clots suggests impending thrombosis and is an indication for further access evaluation (measurement of access flow, imaging by ultrasound or angiography). However, aspiration of clots might also be caused by improper placement of the needle. Thus, puncture should be repeated. Clot aspiration is not an indication for anti-platelet or anti-thrombotic agents.
*5 Prolonged bleeding
• Prolonged bleeding should raise suspicion of high intra-access pressure, outflow (downstream) stenosis or local inflammation. Clinical examination of the site should be performed and previous static or dynamic venous pressure measurements should be reviewed. Measurements of access flow and of static or dynamic venous pressures should be performed during the next treatment as true outflow stenosis means impending thrombosis. However, prolonged bleeding may also be caused by excessive heparinisation of the blood circuit, access laceration during previous cannulation or skin atrophy.
*6 Checks during each HD treatment
• Evidence of access dysfunction can be observed during haemodialysis. Access recirculation, the return of processed blood from the venous needle back up through the arterial needle, can sometimes be observed during dialysis. Recirculation usually occurs when the access flow is lower than the prescribed pump flow (Qb) 6. Recirculation is more commonly seen in A/V fistulas than in PTFE grafts, because grafts tend to clot more rapidly when access flow is less than 600 - 800 ml/min 7 8 9. A/V fistulas, however, usually remain patent even with access flow of less than 400 ml/min.
• Information about possible access dysfunction can be obtained during dialysis by occluding the A/V fistula or graft between the two dialysis needles. In the presence of outflow stenosis, particularly in grafts, venous pressure will rise and arterial pressure may fall. With arterial inflow disease, arterial pressure will fall while venous will remain the same during inter-needle compression 10.
*7 Increasing venous line pressures
• Elevated dynamic or static venous pressures (see *13) are indicative of possible venous outflow stenosis in grafts. However, McDougal et al. found no significant difference in the venous pressure in failed grafts or in grafts that did not fail at blood flow rates of 200 ml/min and 400 ml/min. They concluded, that single static or dynamic venous pressure monitoring was not predictive of graft failure 11.
• Dynamic venous pressure is measured during the first five minutes of the dialysis session (for procedure: see appendix). Dynamic venous pressure repeatedly >150 mm Hg with 16-gauge needle at 200 - 225 ml/min-blood flow may be indicative of possible venous outflow stenosis 12. Thus monthly access monitoring should be anticipated.
• Dynamic pressures are less useful in A/V fistulas than in grafts. Due to their multiplicity of run-off veins, the venous pressure may not rise with outflow stenosis. Upper arm A/V fistulas behave more like PTFE grafts because of their singular venous outflow.
*8 Decreasing arterial line pressures
• Decreasing pre-pump arterial pressure (= high negative pressure) over time (weeks) is another indication of decreased access flow and access dysfunction. Although this can be inferred by noting a collapsed arterial pillow, the measurement of pre-pump arterial pressure is preferable. When the pre-pump arterial pressure is < 150 to – 250 mm Hg, the blood pump is no longer able to deliver the prescribed Qb 13. This is frequently an indication of inflow disease in A/V fistulas and PTFE grafts. These high negative pre-pump pressures also correlate with development of haemolysis 14. If the negative arterial pressure is below –150 mm Hg, the dangerously high negative aspirating pressure inside the vascular access might also damage the access wall.
*9 Checks each month in grafts and in A/V fistulas with dysfunction
• Regardless of the mode of monitoring, most investigators have found significant and cost effective improvements of up to 70 % in cumulative access patency, when prospective monitoring techniques are combined with early correction of stenosis using either angioplasty or surgical revision 12 15 16 17 18 19. Thus these monitoring regimens also help to reduce costs as maintenance of vascular access accounts for 17 – 25 % of all hospitalisations in dialysis patients in the USA 16.
• Screening procedures include physical examination 2, measurement of vascular access flow (Qa), of recirculation, of static and dynamic venous pressure (VP) and colour-coded duplex ultrasound. Monitoring of venous pressure, blood flow or both and intervention in cases of stenoses reduces thrombosis rates to < 0.5 events per patient year, i.e. below the quality recommendation of the DOQI guideline 16 20.
• It is recommended that each unit uses one or more monthly screening tests to monitor vascular access performance in order to predict complications. Unfortunately, the prediction of impending vascular access thrombosis, using one or a combination of simple tests, performed whenever possible during the regular dialysis treatment, is still not perfect. The predictive power is poorer when using a single measurement instead of a trend analysis, or a single technique instead of a combination of parameters 11. Vascular access flow is probably the most useful and reliable parameter available, especially if performed serially 21.
* 10 Dialysis dose
• Low delivered dialysis dose might be indicative of either insufficient blood flow or high recirculation, possibly due to stenosis. However, recirculation by itself is a relative insensitive parameter, detecting stenosis very late 22.
*11 Vascular access flow Qa
Direct measurement of access blood flow (Qa)
• Qa determination is considered the best method to monitor vascular access function and predict vascular access failure 17 21 23 24. Moreover, trend analysis of serial Qa determinations are more effective than a single measurement.
• Qa can be determined by Doppler ultrasound 15 or dilution methods (e.g. haematocrit dilution, thermodilution, conductivity (ionic) dilution) 25 26 27 28 (see appendix).
• It is still controversial whether Qa should be measured only at the beginning of the dialysis session, with stable blood pressure conditions, as vascular access flow may vary over the course of the treatment especially in patients with haemodynamic instability. While Rehman et al. recommend measurement of Qa only during the first 90 minutes of the dialysis session 29, Agharazii et al. found only a non-significant drop in Qa during the haemodialysis session of fifty patients30.
• Angiography should be performed when access flow falls below certain cut-off values. However, it still remains unknown which cut-off value should be used. Cut-off values differ from 500 to 600 ml/min in native A/V fistulas and 650 to 800 ml/min in grafts, or a reduction of Qa of 20-25 % 17 18 31. These differences were to some extent due to using different measurement techniques.
• Neyra et al. found a high risk of thrombosis if Qa decreased by 15 %. Reduction in access blood flow of 35 % increased the risk of thrombosis almost 14-fold compared to accesses with no decrease in blood flow 23.
• However, other authors could not confirm the benefits of a vascular access surveillance program. Paulson et al. found no association between future thrombosis and Qa, or decrease in Qa 32 33. Neither static nor dynamic venous pressure predicted stenosis. Qa was only a poor predictor of stenosis 11. While Wang could find an association between low Qa (< 500 mL/min) and an elevated risk of thrombosis in grafts, this association was not found with A/V fistulas 34.
• Lumsden et al. used duplex ultrasound to detect stenosis > 50 % in ePTFE grafts. Considering the high costs of prophylactic surveillance with duplex ultrasound and the absence of improved patency rates found in their study, Lumsden et al. questioned the use of this technique in grafts 35.
*12 Measurement of recirculation
• Recirculation is less popular as a screening technique since it detects stenosis relatively late at a critical stage. Moreover, its predictive value is especially poor in grafts 36. It is more useful in A/V fistulas provided that the two needles are placed in a segment of the fistula vein without side branches and collaterals (see *13) (for performance of measurement and formula, see appendix).
*13 Static intra-access venous pressure (SVP)
• Static venous pressure is less accurate in A/V fistulas than in grafts, as in the former, collateral vessels might be present or develop, preventing a marked increase in venous pressure 36 (for procedure and formula see appendix). Static venous pressure should be measured every two weeks in grafts.
• However, even in grafts, static venous pressure is not an optimal screening method to predict future thrombosis. For a threshold of SPRv > 0.4 in grafts, sensitivity and specificity of the predictive value for thrombosis within one month was 73 % and 47 %, respectively. With a threshold of SPRv > 0.5, it was 48 % and 75 %, respectively 37.
• A guide for interpreting the pressures measured, according to the DOQI-guidelines, is given in table 3.
Table 3:
Interpretation of the normalized, arterial and venous static intra-access pressures
|
|
Graft - |
A/V fistula - |
||
|
|
SPRa |
SPRv |
SPRa |
SPRv |
|
Normal |
0.35-0.74 |
0.15-0.49 |
0.13-0.43 |
0.08-0.34 |
|
Venous outlet Stenosis |
> 0.75 or |
> 0.5 |
> 0.43 or |
> 0.35 |
|
Intra-access Stenosis |
> 0.75 and |
< 0.5 |
> 0.43 and |
< 0.35 |
|
Arterial inflow Stenosis |
< 0.3 |
- |
< 0.13 |
- |
SPRa = ratio of normalised arterial static intra-access pressure to mean arterial pressure (see appendix)
SPRv = ratio of normalised venous static intra-access pressure to mean arterial pressure (see appendix)
*14 Summary of access monitoring
• Prospective surveillance of grafts to look for stenosis, malfunction and pre-emptive correction probably improves patency rates and life expectancy of the access. In the absence of a gold standard method with high sensitivity and specificity for surveillance and prevention of vascular access failure, each unit should establish its own quality assurance program to monitor prospectively their vascular accesses, combining more than one isolated test (see appendix 2 for proposal of check box).
• There is no evidence that surveillance of autogenous AV fistulas is cost effective.
*15 Appropriate preparation of cannulation site
• In order to reduce the risk of infection and bacteraemia, the cannulation site should be cautiously prepared, according to standards (e.g. as given in the European Best Practice Guidelines - Haemodialysis) 38.
• The needle site should be palpated before preparation of the skin. After cleaning the skin with warm water and soap, chlorhexidine should be applied to the skin in the region of cannulation, as it is considered to be a better antiseptic than povidone iodine. If povidone iodine is used, it should be applied for at least 2 - 3 minutes to allow it to dry. Alcohol (70 %) can be used in patient who have side-effects with the other antiseptics. It should be applied for at least one minute. Wearing new gloves for every patient is mandatory.
• In using the buttonhole technique (see *16 below), puncturing through the scab should be avoided, as this may introduce germs into the access and also cause the needle to puncture and destroy the sides of the tunnel. The scab has to be removed by a different, sterile cannula, before the access can be punctured.
*16 Rotating sites (“rope ladder”)
• The basic knowledge of the types of repeated cannulation were developed by Krönung 39. In most patients with A/V fistulas the “rope ladder“ technique is the favoured technique. In A/V fistulas, the rotating cannulation sites cause a regular, but moderate dilatation of the vein.
• Cannulation of only a small area of the access ("area cannulation") by gradually weakening the access wall causes aneurysms in A/V fistulas and pseudoaneurysms in grafts, thus destroying the latter. Therefore, this technique should be abandoned.
*17 Buttonhole technique
• The “buttonhole“ technique, which has been used exclusively in A/V fistulas, is still discussed controversially. In this technique, also called “constant-site” method, the vascular access is always punctured at the same place, with the same angle, using the same “tunnel” for the cannula. Over time, the tissue around the needle becomes scared, thus directing the needle to the vascular access. It seems to be a good alternative for experienced hands 40 41.
• The buttonhole technique is recommended for self-cannulation. Dull needles have been developed, which reduce the risk of damaging the scared tunnel while cannulating the access. These dull needles can be used after the tunnel has been created. With the buttonhole technique, less bleeding, less pain and less cannulation failure have been reported (see appendix for details of performance).
*18 Mupirocin treatment intranasal
• Patients with nasal carriage of S. aureus are at an increased risk of bacteraemia caused by S. aureus. The infection rate can be reduced fourfold by the eradication of S. aureus by nasal mupirocin treatment 42.
• The European Best Practice Guidelines recommend screening all patients with a past history of S. aureus infection and to consider intervention 38.