Determining the intra- and inter-observer reliability of screening tools used in sports injury research

Abstract

Sports injury etiological studies explore the relationships between potential injury risk factors and injury outcomes. The ability of such studies to clearly identify intrinsic risk factors for sports injury depends on the accuracy of their measurement. Measurements need to be reproducible over time and repeatable by different observers, as well as within a given individual. The importance of the reliability of pre-participation screening protocols and other clinical assessment tools has been identified in a number of published studies. However, a review of these studies indicates that a variety of statistical techniques have been used to calculate intra- and inter-observer reliability. While the intra-class correlation coefficient (ICC) is the most often cited measure, a range of statistical approaches to estimating ICCs have been used. It is therefore difficult to determine which statistical method is most appropriate in the context of measuring intrinsic risk factors in sports injury research. This paper summarises a statistical method for the concurrent assessment of intra- and inter-observer reliability and presents an argument for why this approach should be adopted by sports injury researchers using screening protocols that collect continuous data.