Comparing inter- and intraobserver reliability between two-dimensional and three-dimensional measurements in the tibial component position of unicompartmental knee arthroplasty

Abstract

In unicompartmental knee arthroplasty (UKA), the tibial component has a small coronal plane width, the tibia and tibial component rotations are mismatched, and the large tibial component posterior tilt may make accurate measurements of component positions difficult in radiography. The study aimed to assess the intra- and interobserver reliabilities of radiographic (2D) and 3D computed tomography (3D-CT) measurements and to determine the minimum detectable change (95% confidence level, MDC95) in the tibial component position measurements in UKA. The study included 23 females and 7 males. Two surgeons measured the tibial component position. Intraclass and interclass correlation coefficients (ICC) were calculated to obtain reliability, and Bland–Altman analysis was performed to assess systematic errors. The MDC95 was calculated according to MDC95 = standard error of measurement × 1.96 × √2. In the 2D and 3D- CT measurements, intraobserver reliability for coronal and sagittal positions of the tibial component were sufficiently reliable, where ICCs were >0.8. In the coronal plane, the ICCs for interobserver reliability were lower in 2D (ICC, 0.5-0.7) than in 3D-CT (ICC > 0.9). Bland-Altman plots showed systematic bias in sagittal alignment in the 2D assessment. In the 3D assessment of intra- and interobserver reliability, the MDC95 of the coronal, sagittal, and axial planes was <2°. In the 2D intra- and interobserver reliability, the MDC95 of the coronal and sagittal planes was >2°. The 2D measurement had a risk of misidentifying the tibial component position in UKA.