Predicting computed tomography angiography-measured arteriovenous fistula anastomosis angles using intraoperative non-contrast measurements: a prospective validation study.

BACKGROUND: The anastomotic angle is a modifiable determinant of arteriovenous fistula (AVF) hemodynamics, maturation, and patency. Postoperative assessment typically relies on computed tomography angiography (CTA); however, factors such as contrast exposure, radiation, and cost limit its utility for repeated evaluations. In this study, we derived a correction factor for CTA-measured anastomotic angles. The primary aim was to validate the measurement method, rather than to establish a direct link between the angle and clinical outcomes.

METHODS: This study collected intraoperative measurements from patients undergoing AVF surgeries and used these data to construct a predictive model for CTA-measured anastomosis angles. The model was validated using 20% of the data, with the remaining data used for model fitting. Bootstrap resampling (1,000 samples) was employed to estimate the mean squared error (MSE) and confidence intervals.

RESULTS: The Bland-Altman plot indicated a mean difference of -16.90 between CTA and intraoperative measurements, with the majority of data points falling within the 95% limits of agreement. The Intraclass Correlation Coefficient was 0.79, suggesting acceptable reliability despite notable bias, while the Concordance Correlation Coefficient was 0.45, reflecting the impact of bias. Bootstrap validation yielded a mean MSE of 24.58 with a standard deviation of 15.17. The 95% percentile confidence interval was [2.66, 51.85].

CONCLUSION: This study provides a preliminary adjustment estimate to bridge the gap between intraoperative and CTA-measured angles. While not a fully developed predictive framework, this correction factor offers a practical, radiation-free method for surgeons to approximate the final anastomotic configuration.