Do standardization of the procedure reduce measurement variability of the sonographic anterior drawer test of the ankle?

Abstract

In quantitative ankle stress sonography, different examiners use different techniques, which may cause measurement variability. This study aimed to clarify whether standardizing stress sonography techniques reduces variability in the quantitative measurement of anterior talofibular ligament length change. Fourteen examiners with a mean ultrasound experience of 8.7 years participated in this study. Each examiner performed stress ultrasonography of the ankle using their preferred method on one patient with an intact anterior talofibular ligament (Patient 1) and on two patients with chronic ankle instability (Patient 2 and 3). Changes in the ligament length between the resting and stressed positions were determined. A consensus meeting was then conducted to standardize the sonographic technique, which was used by the examiners during a repeat stress sonography on the same patients. The variance and measured values were compared between the preferred and standardized techniques using F-tests and paired t-tests, respectively. At a consensus meeting, a sonographic technique in which the examiner pushed the lower leg posteriorly against the fixed foot was adopted as the standardized technique. In Patient 1, the change in the anterior talofibular ligament length was 0.4 (range,-2.3-1.3) mm and 0.6 (-0.6-1.7) mm using the preferred and standardized techniques, respectively, with no significant difference in the variance (P=0.51) or the measured value (P=0.52). The length changes in Patient 2 were 2.0 (0.3-4.4) mm and 1.7 (-0.9-3.8) mm using the preferred and standardized techniques, respectively. In Patient 3, the length changes were 1.4 (-2.7-7.1) mm and 0.7 (-2.0-2.3) mm. There were no significant differences between the techniques in either patient group. Variability in the quantitative measurement of ankle stress sonography was not reduced despite the standardization of the technique among examiners. Hence, comparing the measured values between different examiners should be avoided.