Access flow measurement as a predictor of hemodialysis graft thrombosis: making clinical decisions.

Since the introduction of dilution methods for measurement of vascular access blood flow during hemodialysis, more than 170 publications addressing the accuracy, prognostic value, and economic impact of the technology have been presented. Recently researchers (Paulson et al.) have raised concerns about the accuracy of access flow measurements in predicting thrombosis. Our first objective was to address the inadequacies of the analysis by these authors. The second objective was to apply a statistically accepted three-step approach for clinical decision making to assess the utility of access flow surveillance (similar to the K/DOQI guidelines) in the prediction of thrombosis. These steps included 1) estimation of treatment thresholds based on harm-benefit analysis of fistulography-angioplasty versus thrombosis, 2) estimation of prior probability of thrombosis based on patient demographic and clinical characteristics, and 3) application of Bayes' theorem to evaluate whether flow test results provided information that could move patients across the treatment threshold, thus discriminating between patients who should be referred for fistulography-angioplasty and those who should not. These data and an analysis of recent publications show that the implementation of an access flow surveillance program decreases thrombosis rates in hemodialysis units and can significantly reduce the costs associated with hemodialysis access maintenance. We conclude that access flow monitoring (K/DOQI flow thresholds) is useful in the clinical decision-making process for thrombosis prediction across a wide range of demographic categories.