Personalized pain management: The relationship between clinical relevance and reliability of measurements

Abstract

Reliability is a topic in health science in which a critical appraisal of the magnitudes of the measurements is often left aside to favour a formulaic analysis. Furthermore, the relationship between clinical relevance and reliability of measurements is often overlooked. In this context, the aim of the present article is to provide an overview of the design and analysis of reliability studies, the interpretation of the reliability of measurements and its relationship to clinical significance in the context of pain research and management. The article is divided in two sections: the first section contains a step-by-step guide with simple and straightforward recommendations for the design and analysis of a reliability study, with a relevant example involving a commonly used assessment measure in pain research. The second section provides deeper insight about the interpretation of the results of a reliability study and the association between the reliability of measurements and their experimental and clinical relevance. Significance: Reliability studies quantify the measurement error in experimental or clinical setups and should be interpreted as a continuous outcome. The assessment of measurement error is useful to design and interpret future experimental studies and clinical interventions. Reliability and clinical relevance are inextricably linked, as measurement error should be considered in the interpretation of minimal detectable change and minimal clinically important differences.