Inter- and intra-rater reliability for measurement of range of motion in joints included in three hypermobility assessment methods

Abstract

Background Comparisons across studies of generalized joint hypermobility are often difficult since there are several classification methods and methodological differences in the performance exist. The Beighton score is most commonly used and has been tested for inter- and intra-rater reliability. The Contompasis score and the Hospital del Mar criteria have not yet been evaluated for reliability. The aim of this study was to investigate the inter- and intra-rater reliability for measurements of range of motion in joints included in these three hypermobility assessment methods using a structured protocol.

Methods The study was planned in accordance with guidelines for reporting reliability studies. Healthy adults were consecutively recruited (49 for inter- and 29 for intra-rater assessments). Intra-class correlations, two-way random effects model, (ICC 2.1) with 95% confidence intervals, standard error of measurement, percentage of agreement, Cohen’s Kappa (κ) and prevalence-adjusted bias-adjusted kappa were calculated for single-joint measured in degrees and for total scores.

Results The inter- and intra-rater reliability in total scores were ICC 2.1: 0.72–0.82 and 0.76–0.86 and for single-joint measurements in degrees 0.44–0.91 and 0.44–0.90, respectively. The difference between ratings was within 5 degrees in all but one joint. Standard error of measurement ranged from 1.0 to 6.9 degrees. The inter- and intra-rater reliability for prevalence of positive hypermobility findings the Cohen’s κ for total scores were 0.54–0.78 and 0.27–0.78 and in single joints 0.21–1.00 and 0.19–1.00, respectively. The prevalence- and bias adjusted Cohen’s κ, increased all but two values.

Conclusions Following a structured protocol, the inter- and intra-rater reliability was good-to-excellent for total scores and in all but two single joints, measured in degrees. The inter- and intra-rater reliability for prevalence of positive hypermobility findings was fair-to-almost perfect for total scores and slight-to-almost-perfect in single joints.

By using a structured protocol, we attempted to standardize the assessment of range of motion in clinical and in research settings. This standardization could be helpful in the first part of the process of standardizing the tests thus avoiding that assessment of GJH is based on chance.