Observer reliability in ankle and calcaneocuboid stress radiography.

Abstract

Stress radiographic measurements play an important role in assessing the degree of joint instability in scientific investigations and for decision making in treatment. However, their validity and reliability are still a matter of intensive debate. There is no difference regarding interobserver and intraobserver reliability with respect to anterior talar drawer, talar tilt, and calcaneocuboid stress radiographs. Cohort study (diagnosis); Level of evidence, 4. Eighty-nine anterior talar drawer, 89 talar tilt, and 76 calcaneocuboid stress radiographs were selected. Analyses for anterior talar drawer (1 measurement technique), talar tilt (1 measurement technique), and lateral calcaneocuboid instability (4 measurement techniques) were performed by 4 independent raters. One rater repeated the measurements after 1 month. Intraclass and interclass correlation coefficients (ICCs) with calculated confidence intervals assessed intratester and intertester reliability of each measure. Ankle stress radiographic interobserver agreement was ICC = 0.73 to 0.97 for anterior talar drawer test and ICC = 0.78 to 0.97 for talar tilt. Interobserver reliability for calcaneocuboid angle measurement methods was lower (ICC = 0.35-0.91) than for the calcaneocuboid joint-space distance measurements (ICC = 0.81-0.95). Intraobserver ICC varied between 0.78 and 0.97 for ankle stress testing and was 0.67 to 0.94 for calcaneocuboid stress radiography, respectively. Ankle stress radiographic measurements were proven to be reliable. Insufficient reproducibility was found for angular calcaneocuboid stress radiography measurements, while lateral calcaneocuboid joint-space distances offered accurate reliability. Measurement errors can be avoided using standardized stress radiography and measurement techniques with proven reliability.