Comparative Evaluation of UAB and NKF-KDOQI Criteria for Predicting Unassisted Arteriovenous Fistula Maturation Using Postoperative Ultrasound Measurements

Background: This study aimed to compare the predictive performance of the University of Alabama at Birmingham (UAB) criteria with the NKF-KDOQI guidelines in forecasting unassisted arteriovenous fistula (AVF) maturation. Additionally, we sought to evaluate how the predictive accuracy of these criteria varies between upper arm AVFs and forearm AVFs. Methods: We retrospectively analyzed the age, gender, BMI, and comorbidities of chronic kidney disease patients who underwent AVF creation and were monitored from January 2017 to March 2021 at a single medical center. Additionally, we assessed the types of AVFs, the success rate of the first cannulation, and postoperative AVF ultrasound measurements. Results: Among the 560 patients analyzed, 68.6% experienced unassisted AVF maturation. Specifically, upper arm AVFs had a higher unassisted maturation rate (70.2%) compared to forearm AVFs (66.8%). The NKF-KDOQI guidelines had a higher positive predictive value for overall unassisted AVF maturation (0.85) than the UAB criteria (0.77), while UAB showed a better negative predictive value (0.57). Receiver Operating Characteristic Curve analysis revealed that UAB (AUC = 0.643, 95% CI 0.602–0.684) outperformed NKF-KDOQI (AUC = 0.56, 95% CI 0.532–0.589) in predicting overall AVF maturation. For forearm AVFs, UAB (AUC = 0.654, 95% CI 0.593–0.715) also demonstrated superior predictive ability compared to NKF-KDOQI (AUC = 0.517, 95% CI 0.494–0.540), which was close to random chance. Conclusion: In conclusion, the UAB criteria proved superior in predicting the overall unassisted maturation of AVFs compared to the NKF-KDOQI guidelines. This was especially evident for forearm AVFs, where the predictive ability of the NKF-KDOQI criteria for unassisted maturation was nearly random.ROC curve : Forearm AVF