Lucie Risch, Monique Wochatz, Janin Messerschmidt, Tilman Engel, Frank Mayer, Michael Cassel

• The reliability of quantifying intratendinous vascularization by high-sensitivity Doppler ultrasound advanced dynamic flow has not been examined yet. Therefore, this study aimed to investigate the intraobserver and interobserver reliability of evaluating Achilles tendon vascularization by advanced dynamic flow using established scoring systems. Methods-Three investigators evaluated vascularization in 67 recordings in a test-retest design, applying the Ohberg score, a modified Ohberg score, and a counting score. Intraobserver and interobserver agreement for the Ohberg score and modified Ohberg score was analyzed by the Cohen kappa and Fleiss kappa coefficients (absolute), Kendall tau b coefficient, and Kendall coefficient of concordance (W; relative). The reliability of the counting score was analyzed by intraclass correlation coefficients (ICC) 2.1 and 3.1, the standard error of measurement (SEM), and Bland-Altman analysis (bias and limits of agreement [LoA]). Results-Intraobserver and interobserver agreement (absolute/relative)The reliability of quantifying intratendinous vascularization by high-sensitivity Doppler ultrasound advanced dynamic flow has not been examined yet. Therefore, this study aimed to investigate the intraobserver and interobserver reliability of evaluating Achilles tendon vascularization by advanced dynamic flow using established scoring systems. Methods-Three investigators evaluated vascularization in 67 recordings in a test-retest design, applying the Ohberg score, a modified Ohberg score, and a counting score. Intraobserver and interobserver agreement for the Ohberg score and modified Ohberg score was analyzed by the Cohen kappa and Fleiss kappa coefficients (absolute), Kendall tau b coefficient, and Kendall coefficient of concordance (W; relative). The reliability of the counting score was analyzed by intraclass correlation coefficients (ICC) 2.1 and 3.1, the standard error of measurement (SEM), and Bland-Altman analysis (bias and limits of agreement [LoA]). Results-Intraobserver and interobserver agreement (absolute/relative) ranged from 0.61 to 0.87/0.87 to 0.95 and 0.11 to 0.66/0.76 to 0.89 for the Ohberg score and from 0.81 to 0.87/0.92 to 0.95 and 0.64 to 0.80/0.88 to 0.93 for the modified Ohberg score, respectively. The counting score revealed an intraobserver ICC of 0.94 to 0.97 (SEM, 1.0-1.5; bias, -1; and LoA, 3-4 vessels). The interobserver ICC for the counting score ranged from 0.91 to 0.98 (SEM, 1.0-1.9; bias, 0; and LoA, 3-5 vessels). Conclusions-The modified Ohberg score and counting score showed excellent reliability and seem convenient for research and clinical practice. The Ohberg score revealed decent intraobserver but unexpected low interobserver reliability and therefore cannot be recommended.…