Oscillometric pulse wave analysis for detecting low flow arteriovenous fistula.

BACKGROUND: Pulse wave analysis may be useful to assess fistula function. We aimed to prospectively evaluate if convenient oscillometric devices are applicable to detect flow below 500 ml/min in a real life clinical setting.

METHODS: Pulse waves were recorded ambilaterally with the vicorder® device at the brachial artery in 53 patients on haemodialysis with native fistula. Primary variables consisted of the mean slope between the systolic maximum and the diacrotic notch (Slope2), the sum of the mean slopes in the four characteristic sections of pulse waves (Slope∑) and the amplitude of relative volumetric change in the measuring cuff at the upper arm (AMP). Fistula flow was measured with the use of duplex sonography using a standardized approach.

RESULTS: Parameter values above or below the median indicated measurement at the non-fistula side, with sensitivities/specificities of 0.79/0.79 (p < 0.001) for Slope 2, 0.64/0.64 (p = 0.003) for Slope∑ and 0.81/0.81 (p < 0.001) for AMP if measurements at the fistula and non-fistula arm were considered. ROC-analyses of parameter values measured at the fistula to detect low flow demonstrated AUCs (with CI) of 0.652 (0.437-0.866, p = 0.167) for Slope2, 0.732 (0.566-0.899, p = 0.006) for Slope∑ and 0.775 (0.56-0.991, p = 0.012) for AMP. The point with maximal youden's index was regarded as optimal cut-off, which corresponded to sensitivities and specificities of 0.8/0.56 for slope2, 0.86/ 0.56 for Slope∑ and 0.93/0.78 for AMP.

CONCLUSION: Functional surveillance with oscillometry is a promising clinical application to detect a low fistula flow. Among all investigated pulse wave parameters AMP revealed the highest diagnostic accuracy.