Abstract
BACKGROUND: Shear wave elastography (SWE) is utilized to detect the stiffness of the coracohumeral ligament (CHL) for the assessment of a frozen shoulder. However, the consistency of shoulder joint posture during the detection process has not been adequately established. This study aims to evaluate the reliability of various commonly employed shoulder postures in measuring CHL stiffness. METHODS: SWE was used to assess CHL stiffness in 55 patients with phase I frozen shoulder by two independent observers. Two measurements were taken, with a 1-week interval, at three shoulder positions: neutral, 30° external rotation, and 60° external rotation. The intra- and inter-observer reliability of CHL stiffness measurements were analyzed using a two-way randomized model with an intragroup correlation coefficient. RESULTS: Significant differences between intra- and inter-observer measurements of CHL stiffness in the neutral position were observed (P < 0.05). However, no significant differences were found at 30° and 60° of external rotation (P > 0.05). Furthermore, poor intra- and inter-observer reliability of CHL stiffness measurements was established in the neutral position intraclass correlation coefficient (ICC < 0.40), moderate reliability at 30° of external rotation (ICC < 0.40, <.75), and excellent reliability at 60° of external rotation (ICC > 0.75). In addition, the Bland-Altman analyses indicated that the mean differences between intra- and inter-observer measurements of CHL stiffness at 60° of external rotation were closest to the zero scale. CONCLUSION: The reliability of SWE in assessing the stiffness of the CHL in patients with phase I frozen shoulder at 60° external rotation is highly satisfactory.