Validation of a Score for Shoulder Function Evaluation Based on Movement Analysis

Background Measurement of shoulder function is a controversial issue. There is a great variety of measurement tools but none of them has been universally accepted. There is therefore a need to develop extensively validated and convenient measurement tools.

Embedded computerized movement analysis can potentially meet these requirements for measurement of shoulder function. Ambulatory measurement devices allow application in various clinical conditions, display adequate precision and accuracy, and are considerably more straightforward than laboratory-based systems.

Using a Physilog ® II embedded system, Coley (2007) developed a relatively simple score of shoulder function (P Score). The method is based on arm power measurement by three-dimensional accelerometers and gyroscopes during seven consecutive shoulder movements. It demonstrated reliability, responsiveness and criterion-based validity. However, additional knowledge and technological progress could now contribute to further simplification of the.

A secondary analysis of Coley's study data based on principal component analysis and multiple regressions highlighted that a procedure including only two selected movements produces comparable results to P Score. Moreover, the development of wireless systems considerably simplifies set up. Consequently, simpler but equivalent measurement procedure can now be considered.

However some important issues have to be clarified before an extensive validation study can be undertaken. The simplicity of the testing procedure allows test replication. However, the number of movement replications needed to obtain a reliable outcome is presently unknown. Relevance of testing procedure and study feasibility have also to be evaluated. A pilot study is needed to clarify these issues.

Aim The aim of this pilot study is to determine the number of movement replications needed to obtain a reliable result using a simplified cinematic shoulder measurement procedure as well as to evaluate testing procedure and study protocol.

Methods Measurement will be carried out with four groups of patients presenting with frequent shoulder conditions (rotator cuff condition, shoulder instability, diaphyseal or subcapital humerus fracture, frozen shoulder). Measurement procedure includes two consecutive measurements, alternatively conducted by two evaluators. Currently used functional questionnaires will be completed at both stages.

Statistical analysis will address outcome variability according to number of replications and reproducibility.