Measurement properties of the Swedish version of the anterior cruciate ligament return to sport after injury scale (ACL-RSI): A Rasch analysis.

Abstract

This study aimed to investigate the measurement properties, according to the Rasch Measurement Theory, of the anterior cruciate ligament return to sport after injury scale (ACL-RSI) in patients treated with ACL reconstruction in Sweden. ACL-RSI responses from 1065 patients treated with ACL reconstruction were extracted from a rehabilitation-specific registry. Rasch analyses were performed on ACL-RSI item response data using the RUMM2030plus software. The analyses focused on targeting (person-item threshold distribution), item hierarchy, response category thresholds, overall and individual item and person fit, differential item functioning (DIF), unidimensionality and reliability. The ACL-RSI had compromised fit to the Rasch model, including DIF and malfunctioning response categories. Several items correlate with each other and the presumptions to aggregate item responses into one single score were not met. When accounting for local dependency, the measurement properties of the ACL-RSI improved in terms of model fit and DIF and unidimensionality were supported. The ACL-RSI was found to have signs of multidimensionality and local dependency, that is, the answers to one item are influenced by the answers to other items. As such, researchers should be careful when using the ACL-RSI as one single score to evaluate patients treated with ACL reconstruction, unless local dependency is accounted for in the scoring process. Level III.