Validity, reliability and responsiveness in the assessment of pain

Abstract

This paper examines some methodological issues in the assessment of pain, with particular reference to validity, reliability and responsiveness. These issues are illustrated with reference to various measures of pain, in particular the visual analogue scale (VAS). Pain is a complex experience which has sensory, affective, evaluative, cognitive and behavioural dimensions. While the essentially private nature of the pain experience poses problems for criterion-related validity, establishing the content validity of pain assessments is both feasible and necessary to reflect the multidimensional nature of pain. Prescriptive validity is also of importance in clinical contexts. The reliability of most methods of pain assessment is good, but in the case of measures such as the VAS, reliability may be purchased at the cost of content validity. High levels of reliability depend upon careful instruction and standardisation of application. Responsiveness is the ability of an instrument to detect change, and is important in both clinical practice and research. However, instruments which are theoretically highly responsive may be lacking in reliability in practice. The requirements for a pain assessment may vary considerably between clinical and research contexts, and between acute and chronic presentations; in the latter, there may be a need for more qualitative measures. In conclusion, it is argued that the need for methodological refinement in the measurement of pain must be balanced with the need for assessments that are meaningful for the individual patient.