Reliability of radiological and cosmetic measurements in hallux valgus

Abstract

Purpose: To evaluate the reliability of angular, linear, and sesamoid position measurements on preoperative and postoperative radiographs in hallux valgus (HV), as well as cosmetic foot appearance. Material and Methods: Radiographs and photographs from 100 patients undergoing HV surgery were evaluated by two independent observers. Results: Interobserver and intraobserver agreements for HV angle using coefficient of repeatability measures were 4.4° and 3.7°, respectively. Intraclass correlation coefficient measures within and between agreements were 0.97 for HV angle. For intermetatarsal distance, interobserver and intraobserver values were 0.90 and 0.94 when measuring from the midline of each metatarsal and 0.75 and 0.92 when measuring between cortices. Using the visual analog scale to evaluate esthetic appearance, interobserver and intraobserver agreements were 0.59 and 0.79, respectively. Sesamoid position values were also measured. Interobserver and intraobserver kappa values for preoperative and postoperative evaluations with two established methods (Mann or Smith) were 0.47 and 0.70 or 0.65 and 0.75, respectively. Conclusion: Intraobserver reliability was higher than interobserver for intermetatarsal distance, cosmetics, and sesamoid position. Angular measurements were more accurate than linear. Esthetic evaluation was less reliable than radiographic, except in the case of sesamoid position measurements.