Comparative Analysis of 2 Methods for Radiographic Angle Measurement

Abstract

Objective: Angle measurement is a key element to orthopedist’s decision making, especially to hand surgeons. Digital x-ray studies are becoming more frequent, but many of the software used for this purpose do not allow us to perform these measurements. The aim of this study is to evaluate the effectiveness of a method for angle measurement using Microsoft PowerPoint (pp-method), worldwide popular software, compared with traditional measurement with a goniometer. Materials and Methods: We used 50 normal wrist x-ray studies, anterior-posterior and lateral views, randomly selected from a digital archive. Four observers (two hand surgeons and two fellows) performed one series of measurements of the radial tilt (RT) and volar tilt (VT) angle of the distal radius. These measurements were performed by two different methods: with a traditional goniometer (g-method) and by the method described by J. K. Jones et al using pp-method. Measurements were analyzed, and both methods were compared by an analysis of variance (ANOVA). A significance level of 5% was considered. At the same time, kappa index was calculated to evaluate the agreement between observers in each variable by each method. As the kappa index is calculated for variables in categories, two groups were considered: those above and over the median (23.5° for RT and 10° for VT). Results: According to ANOVA, statistical analysis shows that the difference between methods is not significant ( P value < .71 for RT, < .70 for VT). The 95% confidence interval (CI) for g-method is (22.4°-24.1°) for RT and (8.56°-10.79°) for VT, and for pp-method, it is (22.5°-24.2°) for RT and (8.67°-10.9°) for VT. Variation between observers represents 2% of total variation for RT and 1% for VT. Kappa index for g-method and RT variable (g-RT) is 0.426, for pp-RT method is 0.40, for g-VT method, it is 0.371, and for pp-VT method, it is 0.30. Conclusion: Statistical analysis shows that all the observers have the same behavior against both methods ( P value < .82 for RT and < .85 for VT). However, kappa index shows that the agreement between observers is acceptable and that the differences between them represent, at most, 2% of the total variation. According to the results obtained, and having observed that there is no significant difference between traditional measurement with goniometer and the method we proposed to study, we believe that this is a valid and useful tool for radiologic measurement for those orthopedists who do not have a software designed for this purpose. The advantage of having Microsoft PowerPoint in most of the computers of the world makes this method an even more seductive choice.