New psychometric strategies for an appropriate selection and use of outcome measures in physical and rehabilitation medicine

In order to be useful for their intended purposes, outcome measures (rating scales and questionnaires) must provide information that allows valid inferences and decisions to be made. Basic classical test theory is still widely used in peer-reviewed, indexed journals for validating these tools. Classical test theory methods mainly focus on an instrument's total score, whichis simply asserted as the relevant statistic. But, this approach neglects a series of criteria that need to be considered when 25 F. FranChIgnonI i sur.: new psychometric strategies for an appropriate selection and use of .... evaluating the psychometric properties of a measurement tool, and that can be analysed by rasch analysis. The validation activities performed by ra are numerous; the most important ones are those connected with the analysis of: a. dimensionality; b. functioning of rating scale categories; c. internal construct validity of the measure; d. reliability of the scale, in terms of 'separation' (i.e. the ratio of the true spread of the measures with their measurement error). Thus, ra is being increasingly used in the development and evaluation of clinical tools for health care. The purpose of the present paper is to describe the main features of rasch analysis in assessing outcome measures, and summarize some results of our recent psychometric studieson outcome measures in Physical and rehabilitation Medicine, in order to provide insights for the appropriate selection and use of outcome measures.Physiatrists have a responsibility to ensure that measures used in clinical settings are psychometrically sound, and that they are administered thoughtfully and analysed correctly. The contents of this article can bring the final users to critically inspect each outcome measure and the related literature before adopting it for clinical practice, decision making, or policy development.