Reliability of Computer-Assisted and Manual Measurement Methods for Assessment of Graf Type 1 and Type 2 Hip Sonograms.

Abstract

The purpose of this study was to investigate the interobserver, intraobserver, and intermethod reliability of computer-assisted digital and manual measurements of hip sonograms. Seventy-four hip sonograms were evaluated in this retrospective study. Five evaluators measured digital images and manual paper printouts according to the Graf method (Arch Orthop Trauma Surg 1984; 102:248-255). Interobserver and intraobserver reliability rates were calculated. Reliability criteria were graded on a numeric scale. The interobserver reliability of both computer-based and manual methods for alpha angle measurements was good to excellent, but the interobserver reliability was fair to poor for beta angle measurements. Intraobserver reliability was varied. Alpha angle measurements by both manual and computer-based methods had high concordance with each other, whereas beta angle measurements had low concordance. The intermethod variability did not differ between observers. The alpha angle measurements had high concordance with each other for both manual and computer-based methods, whereas the beta angle measurements had low concordance. This information should be taken into account in clinical practice. Overall, the two measurement methods were reliable and consistent with each other.