In search of an optimal bedside screening program for arteriovenous fistula stenosis

Abstract

Background and objectives: Guidelines recommend systematically screening for stenosis using various methods, but no studies so far have compared all of the options. A prospective blinded study was performed to compare the performance of several bedside tests performed during dialysis in diagnosing angiographically proven >50% fistula stenosis.

Design, setting, participants, & measurements: In an unselected population of 119 hemodialysis patients with mature fistulas, physical examination (PE) was conducted; dynamic and derived static venous pressure (VAPR), blood pump flow/arterial pressure (Qb/AP) ratio, recirculation (R), and access blood flow (Qa) were measured; and angiography was performed.

Results: Angiography identified 59 stenotic fistulas: 43 stenoses were located upstream from the venous needle (inflow stenosis), 12 were located downstream (outflow stenosis), and 4 were located at both sites. The optimal tests for identifying an inflow stenosis were Qa < 650 ml/min and the combination of a positive PE "or" Qa < 650 ml/min (accuracy 80% and 81%, respectively), the latter being preferable because it was more sensitive (85% versus 65%, respectively) for a comparable specificity (79% versus 89%, respectively). The best tests for identifying outflow stenosis were PE and VAPR, with no difference between the two (accuracy 91% and 85%, sensitivity 75% and 81%, specificity 93% and 86%, respectively), the former being preferable because it was more reproducible, easier to perform, and applicable to all fistulas.

Conclusions: This study showed that fistula stenosis can be detected and located during dialysis with a moderate-to-excellent accuracy using PE and Qa measurement as screening procedures.

© 2011 by the American Society of Nephrology

Figures

Figure 1.

Diagnostic performance of the tests for inflow stenosis at ROC curve analysis. The diagnostic performance of the various tests was measured from the AUC [95% CI]. In decreasing order, the AUC for Qa (closed circles) was 0.879 [0.815 to 0.944] (P < 0.001), for PE (open circles) it was 0.726 [0.630 to 0.822] (P < 0.0001), for Qb300/AP (open triangles) it was 0.613 [0.510 to 0.717] (P = 0.040), for VP200 (closed squares) it was 0.588 [0.481 to 0.694] (P = NS), for VP300 (open squares) it was 0.543 [0.436 to 0.650] (P = NS), for VAPR (closed triangles) it was 0.510 [0.403 to 0.617] (P = NS), and for R (closed diamonds) it was 0.453 [0.341 to 0.565] (P = NS). The dashed line at a 45° angle indicates no discriminative capacity (AUC = 0.5).

Figure 2.

Diagnostic performance of the tests for outflow stenosis by ROC curve analysis. The diagnostic performance of the various tests was measured from the AUC [95% CI]. In decreasing order, the AUC for VAPR (closed triangles) was 0.853 [0.723 to 0.983] (P < 0.0001), for PE (open circles) it was 0.841 [0.712 to 0.970] (P < 0.0001), for VP200 (closed squares) it was 0.833 [0.726 to 0.940] (P < 0.0001), for VP300 (open squares) it was 0.765 [0.612 to 0.918] (P = 0.001), for Qb300/AP (open triangles) it was 0.730 [0.582 to 0.879] (P = 0.003), for R (closed diamonds) it was 0.585 [0.421 to 0.749] (P = NS), and for Qa (closed circles) it was 0.531 [0.412 to 0.649] (P = NS). The dashed line at a 45° angle indicates no discriminative capacity (AUC = 0.5).

Figure 3.

Diagnostic performance of PE by different raters in detecting inflow stenosis. The diagnostic performance of PE by the various raters was measured from the AUC [95% CI]. In decreasing order, the AUC for inflow stenosis was 0.826 [0.722 to 0.931] (P < 0.0001) for the senior nephrologist (closed circles), 0.649 [0.504 to 0.794] (P = 0.058) for the junior nephrologist (closed triangles), and 0.647 [0.497 to 0.796] (P = 0.064) for the nephrology fellow (open diamonds). The AUC was significantly greater for the senior nephrologist than for the other raters (P = 0.007), with no difference between the latter. The dashed line at a 45° angle indicates no discriminative capacity (AUC = 0.5).

Figure 4.

Diagnostic performance of PE by different raters in detecting outflow stenosis. The diagnostic performance of PE by the various raters was measured from the AUC [95% CI]. In decreasing order, the AUC for outflow stenosis was 0.926 [0.722 to 0.931] (P < 0.0001) for the nephrology fellow (open diamonds), 0.867 [0.728 to 1.000] (P < 0.0001) for the senior nephrologist (closed circles), and 0.831 [0.668 to 0.994] (P = 0.001) for the junior nephrologist (closed triangles). The AUC for the three raters did not differ significantly. The dashed line at a 45° angle indicates no discriminative capacity (AUC = 0.5).