 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
*
Department of Medicine, Dalhousie University, Halifax, Nova Scotia,
Canada.
Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia,
Canada.
Correspondence to Dr. Kailash Jindal, Room 5076, Dickson Building, 5820 University Avenue, Halifax, Nova Scotia, B3H 1V8, Canada. Phone: 902-473-2262; Fax: 902-473-2675; E-mail: kjindal{at}is.dal.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Qa > 20% from baseline or
AR > 5%. Accesses with stenosis underwent percutaneous angioplasty. After 1
yr, there were 1355 mo of follow-up in 177 patients. There were 44 positive
studies in 40 patients. Qa was <500 ml/min in 36 (82%),
Qa was >20% in 5 (11%), and AR was >5% in 6 (14%). Of
patients with Qa < 500 ml/min, 29 (81%) had stenosis. Only two
patients (40%) with Qa > 20% but Qa > 500
ml/min had stenosis. No patient with AR > 5% had stenosis unless
Qa was also <500 ml/min. Immediate patency rate was 93%
post-PTA. Mean Qa increased from 303 ± 154 ml/min to 602
± 220 ml/min (P < 0.0001), and mean urea reduction ratio
increased from 70.4 ± 8.4% to 74.6 ± 6.5% (P = 0.003)
post-PTA. The results demonstrate that UDT could detect subclinical stenoses
in AVF, and most lesions were amenable to angioplasty. AVF that underwent PTA
delivered higher Qa and urea reduction ratio, and immediate patency
rates were acceptable. Access failure after negative UDT was unusual.
Measuring AR increases the time required to perform UDT but does not improve
utility. Serial measurements of Qa alone may be the best strategy
for screening AVF. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Qa > 20% from baseline
or AR > 5% be investigated with angiography
(10), which is similar to the
recommendations of the National Kidney Foundation Dialysis Outcomes Quality
Initiative (11). The value of
implementing these recommendations is unproved. Evidence suggests that low Qa and high AR both are associated with access failure. However, because AR would not be expected to occur except at very low values of Qa (12), it is unclear whether AR is of any incremental value compared with Qa alone. In addition, performance of UDT requires the presence of a single venous segment of at least 4 cm in length. It is not known what proportion of patients with functional fistulae meet this requirement.
There were three objectives for the current study. The first was to test the clinical utility of UDT in detecting subclinical access stenosis in patients with native vessel AVF and specifically to evaluate the performance of the published guidelines. The second was to compare the clinical utility of Qa and AR measurements in this regard. We hypothesized that measurement of AR would not result in improved diagnostic accuracy compared with Qa alone. The third objective was to determine the proportion of patients whose fistulae were anatomically suitable for the use of UDT.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Study Design
This was a prospective observational cohort study. We began screening
patients with functioning native vessel fistulae in October 1999 as per
published guidelines. Screening studies were performed bimonthly with a
Transonic HD01 monitor (Ithaca, NY) using indicator dilution technology
whereby the ultrasound velocity through blood is the indicator and dilution is
provided by the bolus of normal saline. The validation of this device has been
described elsewhere
(3,5).
All studies were performed by a dedicated team of nurses as part of routine
clinical care; the results were reviewed by a vascular access team composed of
nephrologists and dialysis nurses. Each patient had Qa measured
twice in succession during the same dialysis treatment, and the average value
was recorded. AR was measured once and repeated if the first value was not 0.
All measurements of AR were performed using the Transonic device rather than
urea recirculation. Studies were not performed during the last hour of
hemodialysis or during periods of clinically significant hypotension (at the
discretion of the dialysis unit nurses).
The value of Qa indicating that angiography was required (the
"threshold value") was set at 500 ml/min as per published
guidelines (10). A fall of
Qa in excess of 20% from baseline (Qa > 20%)
and the presence of AR > 5% also were indications to perform angiography.
The attending nephrologist was not routinely notified of the results of UDT
studies; angiography was ordered by the vascular access team when the
specified criteria were met. Patients for whom the attending nephrologist had
ordered a fistulogram because of clinically suspected access dysfunction
(i.e., for any reason except for UDT data) were excluded, as the
focus of this analysis was on cases of subclinical stenosis detected only by
screening.
Fistula stenosis was defined by a >50% reduction in vessel diameter (11) as reported by the attending radiologist. Successful angioplasty was defined as an increase in vessel diameter postprocedure (as reported by the radiologist) in a fistula that could be used for dialysis. Patency was defined as suitability for clinical use; secondary patency included fistulae that had required multiple percutaneous interventions. Immediate and secondary patency were defined as per published guidelines (13). Follow-up continued for 12 mo after the implementation of the screening program. Pre- and postprocedure values for Qa and URR were defined by the routine measurements immediately preceding and following the intervention. Because routine studies were performed bimonthly, the interval between measurement of these parameters and performance of angioplasty was never more than 2 mo and often was substantially shorter.
Statistical Analyses
Positive predictive value (PPV) and negative predictive value (NPV) were
calculated using standard formulas. Paired two-sided t tests were
performed on pre- and postprocedure values for Qa and URR. The
sample correlation coefficient (r) was calculated for pairs of
Qa measurements as an index of reproducibility within a given
treatment. The level of statistical significance was set at P = 0.05.
Results are expressed as mean ± SD or as percentages. Statistical
analysis was performed using the Statistica software package (Statsoft Inc.,
Tulsa, OK).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analysis of the first 100 paired measurements of Qa showed a
high degree of reproducibility within a given dialysis treatment (r =
0.98, P < 0.0001). Using the definitions above, there were 44
positive studies in 40 patients (22.6% of the cohort). Qa was
<500 ml/min in 36 (82%), Qa was >20% in 5 (11%), and
AR was >5% in 6 (14%). Some patients had a positive study by more than one
criterion.
A positive study led to a change of needle placement in one patient; subsequent studies with needle position optimized showed greatly improved Qa, and angiography was not performed. The remaining 43 positive studies were investigated by angiography, and 31 (72%) were found to have angiographically significant stenoses. There were 22 stenoses (71%) within 2 cm of the anastomosis, 12 peripheral stenoses (39%) that were >2 cm distal to the anastomosis, and 3 central stenoses (10%) detected. Some patients had more than one site of stenosis.
Twenty-five of 31 patients (81%) with stenosis had a successful fistuloplasty. One patient with stenosis died, and three other fistulae with stenosis failed before angioplasty could be performed (mean Qa, 250 ± 35 ml/min; all were <300 ml/min). In the remaining two accesses, angioplasty was unsuccessful; one of these patients required insertion of a temporary dialysis catheter as the fistula was no longer usable. The other access continued to be used for dialysis, but neither the calibre of the vessel nor Qa improved postprocedure. Four of the 25 patients required additional interventions to maintain fistula patency. The secondary patency rate for the 27 patients in whom fistuloplasty was attempted was 78.6% at 6 mo. No fistula was lost during diagnostic angiography. Defining successful angioplasty in terms of an increase in Qa rather than by angiographic appearance did not significantly change our results, as 23 of 25 patients with an increased fistula diameter on angiography also had increased Qa postprocedure.
Excluding the patient whose access failed postprocedure, mean Qa increased from 303 ± 154 ml/min preangioplasty to 602 ± 220 ml/min postangioplasty (P < 0.0001). Mean URR also increased from 70.4 ± 8.4% preprocedure to 74.6 ± 6.5% postprocedure (P = 0.003). There were no changes in dialysis prescription postprocedure, and mean Qb was not recorded.
Of patients with Qa < 500 ml/min, 29 of 36 (81%) had stenosis
on angiogram, and 24 (83%) had successful angioplasty. Only two of five
patients (40%) with Qa of >20% (but in whom Qa
> 500 ml/min) had stenosis on angiography; both lesions were amenable to
angioplasty. Neither of two patients with AR > 5% and Qa >
500 ml/min had stenosis, but four of four with AR > 5% and Qa
< 500 ml/min had stenoses amenable to angioplasty. No patient had AR <5%
but >0%.
In three patients, Qa could not be measured despite a single venous segment and AR = 0%. All three underwent angiography and were found to have stenoses amenable to fistuloplasty.
Twelve patients had positive screening studies but negative angiograms.
Seven such patients had Qa < 500 ml/min. Two of these had needle
placement optimized once their anatomy was clarified, with subsequent
improvement in Qa. Three other patients had small-calibre vessels
rather than stenosis, and in the remaining two patients, no explanation for
the false-positive screening study could be identified. Two patients with
negative angiograms had AR > 5% and Qa > 500 ml/min; one had
needling optimized after angiography, with resolution of AR, and in the other,
no explanation was found (AR was never detected subsequently in the access).
Three patients with negative angiography had Qa > 500 ml/min, no
AR, and Qa > 20%. We could not detect a cause for the
fall in Qa observed in these patients, and subsequent Qa
measurements were stable or improved.
Sensitivity, specificity, and true NPV could not be calculated, because not all fistulae underwent angiography. However, among the 109 accesses in which UDT were negative during the first 6 mo of the study, only 2 subsequently failed during the follow-up period. Mean duration of follow-up in these 109 fistulae was 9.7 mo. Therefore, the NPV of UDT for access failure in this setting is approximately 98% (107 of 109) at 10 mo.
The sensitivity and specificity of UDT for stenosis can be estimated by assuming that all fistulae that had negative UDT and did not fail during the study period would have had negative angiograms. On this basis, the three criteria for positive UDT had a combined sensitivity of 94% (31 of 33) and combined specificity of 92% (135 of 147). Given the study design, it was not meaningful to perform receiver operating curve analysis to find the Qa with optimal sensitivity and specificity for stenosis. As shown in Table 1, lowering the Qa threshold at which angiography was performed would have improved PPV. However, it is unclear what effect (if any) this would have had on NPV.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We found that Qa < 500 ml/min has a high PPV for stenosis, which is similar to other work done in native vessel fistulae (6). Diagnostic performance was not improved by measuring AR, which is not surprising because significant AR would not be expected when Qa is >500 ml/min (12,16). In addition, several patients did not have detectable AR despite very low Qa. In all cases, this occurred in association with an intra-access stenosis (between the needles used to cannulate the fistula), which explains the absence of detectable recirculation. Thus, AR does not occur until Qa is significantly reduced and is not acceptably sensitive even at very low Qa. In addition, measuring AR as well as Qa significantly increases the time required to perform UDT. Because the labor costs of UDT may be reduced if they can be performed more quickly, we suggest that only Qa be used to screen for subclinical access dysfunction in AVF.
A threshold value of Qa < 500 ml/min was used in our study, as recommended by published guidelines. All three accesses that failed before fistuloplasty could be performed had Qa < 300 ml/min. Because PPV increases with reductions in the threshold value for Qa, it is possible that a threshold value of 400 ml/min might allow increased PPV without unduly sacrificing NPV. Our data are insufficient to make such a recommendation, and this issue deserves further study.
Seven of 36 patients (19%) had Qa < 500 ml/min but no stenosis on angiogram. In most cases, this was due to needling technique or small native vessels. Other theoretical explanations for this phenomenon include congestive heart failure, arterial insufficiency in the ipsilateral extremity, operator error, or technical measurement error. We were not able to implicate conclusively any of these factors in our patients.
In our study, the Qa parameter was associated with a
lower PPV than Qa alone. Because Qa may vary up to 10%
within a single treatment
(17), a difference of 20%
between screening studies may be insufficiently specific for stenosis. We did
not perform UDT during the last hour of dialysis treatments or during
hypotensive episodes, which are associated with increased variability in
Qa. Although others found that decrements in Qa were
strongly associated with access failure
(18,19),
the great majority of accesses studied were polytetra-fluoroethylene grafts,
and the applicability of this finding to AVF is unclear.
No screening for subclinical access dysfunction was performed in our unit
before this study. This may have resulted in a larger proportion of severe
stenoses among the detected cases, which would be expected to improve the PPV
of Qa compared with Qa. Therefore, further work
should be done to determine the role of Qa, including
determination of the value with optimal diagnostic properties.
Because our study was uncontrolled, we are unable to assess the impact of screening on fistula survival. Although an improvement in Qa and in the degree of stenosis was demonstrated in most patients who underwent angiography, we did not perform follow-up angiograms to rule out recurrent stenosis or vessel recoil. It is of note that 3 of 31 accesses that were found to have stenosis failed before angioplasty could be performed, suggesting that the likelihood of access thrombosis in patients with Qa < 500 ml/min is high. This is in agreement with the available literature, including a previous cohort of native vessel fistulae (6,7,8,9,20).
Low Qa may be a clue to improper needle placement in the absence of access stenosis. In our cohort, 3 of 12 "false positive" studies were due to difficulties with needling of the fistula, which resolved once the problem was identified, permitting higher Qa. In addition, the inability to measure Qa (persistence of the "unusual curve" error message) despite correct technique and a single venous segment was strongly associated with stenosis. In one case, there was an intra-access stenosis (between the dialysis needles), and in another, extensive development of collateral vessels was apparent on angiography. Thus, the inability to measure Qa in this situation may be due to flow of the saline bolus through a collateral vessel, without ever being detected by the second ultrasound probe. Regardless of the mechanism responsible for this finding, nephrologists should consider arranging fistulography in this situation. For purposes of this analysis, we did not consider patients without measurable Qa as cases detected by screening, because they did not meet our a priori definition of a positive study.
We found improved dialysis delivery after percutaneous intervention, despite that access dysfunction was not clinically suspected in these patients. Because dialysis prescription did not change and AR was evident in only a minority of patients, presumably the fistulae were able to deliver higher Qb after successful angioplasty. Unfortunately, we did not record Qb to confirm this hypothesis. Other workers noted an association between low Qa and dialysis delivery but did not remeasure solute clearance after intervention (6).
Both Canadian (10) and American (11) guidelines recommend routine screening of native vessel AVF, with percutaneous or surgical correction if stenoses are found. There is no evidence that this strategy will extend fistula survival. However, this study shows that subclinical stenoses in AVF can be detected and percutaneously corrected, with acceptable immediate and secondary patency rates. In comparison, the high failure rates of stenosed fistulae that do not undergo angioplasty (6,8) suggest but do not prove that screening is beneficial.
Limitations of our study include the lack of a control group and our inability to calculate sensitivity and specificity for these techniques. In addition, our results may not apply to units that currently screen for access dysfunction using other methods such as static or dynamic pressure monitoring or Doppler ultrasound. Finally, we were not able to determine the optimal frequency of screening, because all patients were studied every 2 mo.
In summary, UDT are useful for detecting subclinical stenoses in native vessel AVF, and the majority of these accesses are anatomically suitable for study. Correcting the stenoses detected by UDT screening permitted higher Qa and improved dialysis delivery. Performing angiography when Qa was <500 ml/min or could not be measured was associated with a high PPV, although the ideal threshold value for Qa remains unknown. Because measuring AR did not improve utility and is time consuming, we suggest that AR not be used routinely to screen for access dysfunction in native vessel AVF. Additional work should be done to confirm the diagnostic importance of decrements in Qa and the optimal screening frequency in this population.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. Ponce, A. Mateus, and L. Santos Anatomical correlation of a well-functioning access graft for haemodialysis Nephrol. Dial. Transplant., September 5, 2008; (2008) gfn512v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. R. Polkinghorne, K. K. P. Lau, A. Saunder, R. C. Atkins, and P. G. Kerr Does monthly native arteriovenous fistula blood-flow surveillance detect significant stenosis--a randomized controlled trial Nephrol. Dial. Transplant., September 1, 2006; 21(9): 2498 - 2506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Jindal, C. T. Chan, C. Deziel, D. Hirsch, S. D. Soroka, M. Tonelli, and B. F. Culleton CHAPTER 4: Vascular Access J. Am. Soc. Nephrol., March 1, 2006; 17(3_suppl_1): S16 - S23. [Full Text] [PDF]
|
|
|
|
|
|
B. Manns, M. Tonelli, S. Yilmaz, H. Lee, K. Laupland, S. Klarenbach, V. Radkevich, and B. Murphy Establishment and Maintenance of Vascular Access in Incident Hemodialysis Patients: A Prospective Cost Analysis J. Am. Soc. Nephrol., January 1, 2005; 16(1): 201 - 209. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Tonelli, D. J. Hirsch, C. T. Chan, J. Marryatt, P. Mossop, C. Wile, and K. Jindal Factors associated with access blood flow in native vessel arteriovenous fistulae Nephrol. Dial. Transplant., October 1, 2004; 19(10): 2559 - 2563. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Tessitore, G. Lipari, A. Poli, V. Bedogna, E. Baggio, C. Loschiavo, G. Mansueto, and A. Lupo Can blood flow surveillance and pre-emptive repair of subclinical stenosis prolong the useful life of arteriovenous fistulae? A randomized controlled study Nephrol. Dial. Transplant., September 1, 2004; 19(9): 2325 - 2333. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Tonelli, G. S. Jhangri, D. J. Hirsch, J. Marryatt, P. Mossop, C. Wile, and K. K. Jindal Best Threshold for Diagnosis of Stenosis or Thrombosis within Six Months of Access Flow Measurement in Arteriovenous Fistulae J. Am. Soc. Nephrol., December 1, 2003; 14(12): 3264 - 3269. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Tessitore, G. Mansueto, V. Bedogna, G. Lipari, A. Poli, L. Gammaro, E. Baggio, G. Morana, C. Loschiavo, A. Laudon, et al. A Prospective Controlled Trial on Effect of Percutaneous Transluminal Angioplasty on Functioning Arteriovenous Fistulae Survival J. Am. Soc. Nephrol., June 1, 2003; 14(6): 1623 - 1627. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Tonelli, D. Hirsch, T. W.I. Clark, C. Wile, P. Mossop, J. Marryatt, and K. Jindal Access Flow Monitoring of Patients with Native Vessel Arteriovenous Fistulae and Previous Angioplasty J. Am. Soc. Nephrol., December 1, 2002; 13(12): 2969 - 2973. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |
