Validation of the inverse method for the determination of the access flow with thermodilution

Thermodilution is a widely used method for measuring vascular access flow (Q_A). Among the possibilities of thermodilution, the reverse method (RM) can be beneficial in the execution time, without impact on the dialysis efficacy (K_t). However, it is not a sufficiently studied technique.

Transversal study of 117 arteriovenous fistulas. Two Q_A measurements were taken with the method described by the manufacturer (MR) and another with RM. RM is based on the obtention of an inverted recirculation registry at the beginning of the session and a single subsequent recirculation measurement with the lines in normal position. In the concordance analysis, the Bland-Altman method and Cohen's Kappa index were used.

Very good concordance between MR and RM was evidenced for Q_A below 700ml/min, but it worsens as flow increases. The median variability between the MR measurements (intra-method variability) was 3.4% (-17.13). This value did not differ from the median variability generated between MR and RM (inter-method variability), which was 2% (-14,12) (P=.287). The degree of agreement between the 2 to identify arteriovenous fistulas susceptible to intervention was very good (Kappa=0.834). The time spent using the RM was significantly shorter (P=.000) without evidence of variations in the K_t of the measurement sessions (P=.201).

The thermodilution RM is valid to determine the flow of the vascular access, especially in Q_A lower than 700ml/min, with great time savings, simplification of the procedure and without modifying the dialysis efficiency. The variability between the measurement by MR and RM is similar to that of MR. The concordance between methods in identifying potentially pathological arteriovenous fistulas is very good.