Diagnostic Accuracy of the Mobile Assessment of Varus Thrust Using Nine-axis Inertial Measurement Units

OBJECTIVES: The purpose of this study was to clarify the diagnostic accuracy of the mobile assessment of varus thrust using inertial measurement units (IMUs). METHODS: A total of 80 knees in 49 patients were enrolled in this study. On visual analysis of gait to determine the presence or absence of varus thrust, 23 knees were assigned to the Present group, 17 to the Ambiguous group, and 40 to the Absent group. The peak knee varus angular velocities (PVVs), measured by quantitative gait analysis using nine-axis IMUs, were compared between these three groups. A receiver operating characteristic curve for the relationship between the visual assessment of varus thrust (Present and Ambiguous) and the measured PVV was created, and the cut-off PVV for visualized varus thrust was determined as the highest point for both sensitivity and specificity. RESULTS: The mean PVVs were significantly different between the three groups (Present, 47.7 ± 8.2 degree/s, Ambiguous, 34.1 ± 10.5 degree/s, and Absent, 28.1 ± 8.3 degree/s, respectively, ANOVA P=0.000). The PVV cut-off value for visualized varus thrust was 28.1 degree/s, yielding a sensitivity of 0.957 and a specificity of 0.579. CONCLUSIONS: A PVV <28.1 degree/s is useful for ruling out varus thrust during gait. This quantitative varus thrust assessment method using IMUs has clinical utility as a screening test.