Native arteriovenous (AV) fistulae have better patency and require fewer interventions compared with synthetic grafts, 1,2 so that the National Kidney Foundation Dialysis Outcome Quality Initiative (NKF-DOQI) guidelines of 1997 recommended to increase the placement of native AV fistulae for hemodialysis. 3 As a result of this recommendation, the number of native fistulae significantly increased in the following years. Still, thrombosis remains the main cause of fistula loss, 4 and low access flow is considered the most important factor in this development. 2,3 Thus, access flow measurements should also be used in fistulae to assess fistula function. However, the measurement of access flow is not done without difficulty, and there is an increased interest in practical alternatives. 5–8 One of the under-recognized alternatives in monitoring fistula function is based on the measurement of access recirculation. Because fistulae remain patent even when access flow decreases to a level below the prescribed pump blood flow (i.e., < 300 to 500 ml/min with high efficiency hemodialysis), the occurrence of access recirculation could be used to identify fistulae with insufficient blood flow. 3,4 Apart from the increased risk of thrombosis associated with low access flow, access recirculation causes a considerable reduction in hemodialysis efficiency and therefore also represents an important functional impairment of the peripheral access.
The purpose of this prospective study done with all patients with AV fistulae at one dialysis center was to determine the value of a simple and time efficient recirculation technique to be used with every hemodialysis treatment in identifying fistula function and to refer those with insufficient blood flow and inefficient hemodialysis for revision. Furthermore, we wanted to test whether the success of detecting access recirculation in grafts using a technique that includes effects caused by cardiopulmonary recirculation could be extended to fistulae. 9
Materials and Methods
The study was performed during a 6 month period as approved by the Beth Israel Medical Center Institutional Review Board. All chronic hemodialysis patients with AV fistulae from the same out-patient center participated in this study. Fistulae had been used for at least 1 month. Details of each patient’s access and clinical information were recorded at the beginning of the study. Bicarbonate hemodialysis was performed using Fresenius 2008H machines (FMC-North America, Walnut Creek, CA). All patients received hemodialysis thrice weekly for 3 to 4 hours using high flux polysulfone dialyzers. Extracorporeal blood flow ranged from 300 to 500 ml/min, and dialysate flow was 800 ml/min.
Recirculation was measured using the Blood Temperature Monitor (BTM) (FMC-North America). The accuracy and repeatability of BTM thermodilution has been previously validated. 9,10 In one of these studies done in polytetrafluoroethylene grafts, we also demonstrated that a BTM recirculation larger than 15% was highly significant for access recirculation. 9
Access flow was measured by ultrasound dilution using the Hemodialysis Monitor (HDM01, Transonic Systems Inc., Ithaca, NY). 11 All flow measurements were performed at a constant extracorporeal blood flow of 300 ml/min. As recommended by the manufacturer, ultrafiltration was turned off 3 minutes before the measurement to avoid the effects of hemoconcentration.
Measurement of recirculation was performed within the first hour of each hemodialysis treatment. When two consecutive recirculation measurements were greater than the threshold of 15%, access flow was immediately measured by ultrasound dilution. If access flow was < 500 ml/min, patients were referred for revision. There was no control group because delivery of inefficient dialysis was considered unacceptable. Patients were followed up for outcomes such as thrombosis, placement of a new access, and surgical or angioplastic revision.
Data are presented as mean values ± SD. Differences between groups were assessed by chi-square-test or Student’s t-test. Statistical significance was defined as a probability p < 0.05 to reject the null hypothesis.
At the beginning of the study, 183 adult hemodialysis patients were treated in the study center. Synthetic grafts and dual lumen catheters served as the primary vascular accesses in 35 and 17% of patients, respectively; 88 patients (48%) had a native AV fistula as their primary access, but 8 patients did not complete the 6 month study period because 3 patients died, 3 patients transferred to other dialysis units, and 2 patients underwent renal transplantation. Eighty patients entered the final analysis. Characteristics of patients with AV fistulae that were included in the final analysis compared with patients with synthetic grafts are shown in Table 1. Men, whites, and patients without diabetes were more likely to have fistulae, whereas women, blacks, and patients with diabetes were more likely to have grafts.
Only 10 fistulae were identified with a BTM recirculation above the threshold of 15%. Nine of the 10 fistulae had an access flow of less than 500 ml/min. These fistulae were referred for surgical or angioplastic revision. The other fistula with a recirculation above the threshold had an access flow of 1,550 ml/min and was not referred for revision. Two fistulae thrombosed in spite of a recirculation below the threshold (9.5 and 12.9%, respectively), so the thrombosis rate was 0.05 per fistula-year at risk. Blood flow in these fistulae was 400 and 450 ml/min, respectively. Eleven angioplastic or surgical revisions were performed during the 6 month observation phase, corresponding to an intervention rate of 0.28 interventions per fistula-year at risk. The sensitivity and specificity of a BTM recirculation above the threshold in predicting an intervention was 81.8 and 98.6%, respectively. The false positive rate was 1.5%, the false negative rate was 18.2%, and the odds ratio was 306 (Table 2).
In this study we investigated whether frequent recirculation measurements by thermodilution could be used to identify fistulae with low access blood flows because such fistulae are at an elevated risk of thrombosis and because the delivery of adequate dialysis is severely compromised in the presence of access recirculation. 12,13 Recirculation by BTM thermodilution using a threshold of 15% was highly sensitive (81.8%) and specific (98.6%) in identifying required surgical or angioplastic revisions. Only 2 of 80 fistulae studied during the 6 month period thrombosed without producing the trigger signal for early intervention.
Access flow has been proposed as the best test for fistula function. 3,14 However, the measurement of access flow in native fistulae is not without problems, and there is an increased interest in practical alternatives. 7,8,15,16 It also has been shown that monthly flow measurements do not result in a further reduction of the already very low thrombosis rate in native fistulae. 17 The threshold for critical blood flow is much lower in fistulae than in synthetic grafts and is close to the range of extracorporeal blood flows of 400 to 500 ml/min used with high efficiency hemodialysis. In a previous study, we found that four of six fistulae thrombosed at an access flow of < 500 ml/min. 18 Besarab et al.14 also reported that most fistula only thrombosed when access flow dropped below 500 ml/min. Thus, there is a high probability for access recirculation to develop in fistulae before thrombosis as access flow reaches a critical blood flow. 19,20 The critical access flow thus can be expected to correspond to a critical recirculation above which the risk of fistula thrombosis increases. However, the threshold for a decline in dialysis efficiency is determined by the onset of true access recirculation. The conversion of critical access flow to critical recirculation bears a major technical benefit: it is much easier to measure recirculation than to measure access flow because there is no need to reverse blood lines, and there is no loss in efficient treatment time. When recirculation is measured by BTM thermodilution, the threshold to detect access recirculation is given by a BTM recirculation of 15%, which includes effects caused by cardiopulmonary recirculation as described elsewhere. 9
In a previous study, 63 fistulae were monitored by measurement of access flow at 2 month intervals for a period of 6 months, and thrombotic events were followed, but fistulae with low access flow were not referred for revision. 18 Six thrombotic events occurred in 63 fistulae during the study period, corresponding to an event rate of 0.19 per fistula-year at risk. The incidence of fistula thrombosis was much lower in this study (0.05), probably because fistulae at risk were referred for early revision. However, the difference in thrombosis rate was not significant when compared with the previous study (χ2 = 2.10, p = 0.16). The reduction of fistula thrombosis rate when AV fistulae are referred for revision at the onset of access recirculation remains to be confirmed in an extension of this study. However, the benefit of recirculation measured with every treatment is evident. Failure to deliver the prescribed dose of dialysis is immediately recognized, long before obtaining indirect evidence from monthly urea kinetic measurements.
Access recirculation has been regarded as a poor test for detecting access problems because it occurs after access flow has decreased to the point of impairing dialysis delivery. 14,19 In addition, the interval between recirculation measurements in previous studies may have been too long to reflect the decline in access flow and the onset of access recirculation. The advantage of recirculation measurements obviously increases with measuring frequency. Frequent recirculation measurements not only provide information on needle placement but also yield instant bedside information on access function in routine dialysis. Therefore, with the prevalence of AV fistulae, access recirculation should not be dismissed as a gadget to monitor access function, especially when measurements can be frequently performed.
Although the incidence of thrombosis was greatly reduced in this study, two fistulae thrombosed without obvious access recirculation.
Because access recirculation depends upon the ratio of blood flow to access flow, one has to ask whether these fistulae remained undetected because of low blood flow. However, in both instances, blood flow was between 400 and 450 ml/min on the treatment day before clotting occurred, so that access flow must have been in excess of 400 and 450 ml/min. This points to an important limitation of the approach. The detection of a critical access flow by the measurement of access recirculation requires a blood flow at or above the value of the access flow. If the critical access flow is 400 ml/min, extracorporeal blood flow will have to be larger than 400 ml/min to produce access recirculation. Such high blood flows are only used with high efficiency hemodialysis, and the use of access recirculation to detect accesses at risk will be limited to this treatment mode. However, should the critical access flow be lower than 400 ml/min, this approach could also be extended to a larger dialysis population treated by conventional hemodialysis.
Absence of access recirculation is no absolute guarantee for absence of access problems. One of the reasons is because of the effect of intra-access stenosis. When intra-access stenosis produces a significant resistance to access flow between arterial and venous needle puncture sites, blood flow evades intra-access resistance during hemodialysis and takes a functional bypass through the extracorporeal bloodlines to the downstream section of the access. Even though such a fistula is under high risk for future thrombosis because of intra-access stenosis, there is sufficient blood supply to the inlet of the arterial bloodline, and there is no recirculation during hemodialysis. 20 However, with frequent measurement of access recirculation, and with rotation in the needle puncture sites, the sensitivity of recirculation measurements to detect an intra-access stricture might be increased. On the other hand, because of the more complex anatomy of AV fistulae, measurement of access recirculation occasionally can identify a dysfunctional fistula even with sufficient access flow. 21
One single fistula presented a BTM recirculation above the threshold of 15%; however, access flow determined by ultrasound dilution was 1,550 ml/min. We have previously shown that a combination of very low cardiac output (< 3.5 L/min) with high access flow (> 2.0 L/min) will lead to a BTM recirculation larger than 15% and to false positive results. 9 Although this combination of patient variables is rare and unlikely to occur, a false positive result is no major disadvantage because all positive results can be evaluated by an access flow measurement.
In summary, BTM recirculation is easily performed in everyday dialysis so that access function can be monitored with every hemodialysis treatment.
1. Woods JD, Turenne MN, Strawderman RL, et al: Vascular access survival among incident hemodialysis patients. Am J Kidney Dis 30: 50–57, 1997.
2. Schwab SJ: Vascular access for hemodialysis. Kidney Int 55: 2078–2090, 1999.
3. NKF-DOQI: Clinical practice guidelines for vascular accesses. Am J Kidney Dis 30: S150–S191, 1997.
4. Albers FJ: Causes of hemodialysis access failure. Adv Renal Replace Ther 1: 107–118, 1994.
5. Bonucchi D, D’Amelio A, Capelli G, Albertazzi A: Management of vascular access for dialysis: an Italian survey. Nephrol Dial Transplant 14: 2116–2118, 1999.
6. Ezzahiri R, Lemson MS, Kitslaar PJ, Leunissen KM, Tordoir JH: Haemodialysis vascular access and fistula surveillance methods in the Netherlands. Nephrol Dial Transplant 14: 2110–2115, 1999.
7. Skladany M, Vilkomerson D, Lyons D, Chilipka T, Delamere M, Hollier LH: New, angle-independent, low cost Doppler system to measure blood flow. Am J Surg 176: 179–182, 1998.
8. Steuer RR, Miller DR, Zhang S, Bell DA, Leypoldt JK: Noninvasive transcutaneous determination of access blood flow rate. Kidney Int 60: 284–291, 2001.
9. Wang E, Schneditz D, Kaufman AM, Levin NW: Sensitivity and specificity of detection of access recirculation by thermodilution technique. Nephron 85: 134–141, 2000.
10. Schneditz D, Wang E, Levin NW: Validation of haemodialysis recirculation and access blood flow measured by thermodilution. Nephrol Dial Transplant 14: 376–383, 1999.
11. Krivitsky NM: Theory and validation of access flow measurement by dilution technique during hemodialysis. Kidney Int 48: 244–250, 1995.
12. Golledge J, Smith CJ, Emery J, Farrington K, Thompson HH: Outcome of primary radiocephalic fistula for haemodialysis. Br J Surg 86: 211–216, 1998.
13. Berardinelli L, Vegeto A: Lessons from 494 permanent accesses in 348 hemodialysis patients older than 65 years of age: 29 years of experience. Nephrol Dial Transplant 13: 73–77, 1998.
14. Besarab A, Lubkowski T, Frinak S, Ramanathan S, Escobar F: Detecting vascular access dysfunction. ASAIO J 43: M539–M543, 1997.
15. Krivitiski NM, Depner TA: Development of a method for measuring hemodialysis access flow: From idea to robust technology. Semin Dial 11: 124–130, 1998.
16. Blankestijn PJ, Smits JHM: How to identify the haemodialysis access at risk of thrombosis? Are flow measurements the answer? Nephrol Dial Transplant 14: 1068–1071, 1999.
17. Sands JJ, Kapsick B, Jabyac P: Monthly access flow monitoring decreases thrombosis in PTFE grafts but not in AV fistulas [abstract]. J Am Soc Nephrol 9: 183A, 1998.
18. Wang E, Schneditz D, Nepomuceno C, et al: Predictive value of access blood flow in detecting access thrombosis. ASAIO J 44: M555–M558, 1998.
19. Besarab A, Sherman RA: The relationship of recirculation to access blood flow. Am J Kidney Dis 29: 223–229, 1997.
20. Schneditz D: Recirculation, a seemingly simple concept. Nephrol Dial Transplant 13: 2191–2193, 1998.
21. Krisper P, Aschauer M, Tiesenhausen K, Leitner G, Holzer H, Schneditz D: Access recirculation in a native fistula in spite of a seemingly adequate access flow. Am J Kidney Dis 35: 529–532, 2000.