Measure of experimental pain using Rasch analysis.

Most common instruments used to assess the painfulness of nociceptive stimuli and the perception of such stimuli are ordinal. This property limits arithmetical operations and statistical procedures that can be applied on their numbers. The Rasch methodology provides mathematical procedures for transforming scores on an ordinal scale into measures on an interval scale. The present paper aims at presenting the basics of this methodology by applying it to the measurement of experimentally induced pain. Six blocks of seven CO(2) laser heat stimuli varying in intensity were delivered on the hand of 100 healthy subjects. They rated their pain perception on a three-level verbal rating scale (not painful, slightly painful, painful). One member of the family of Rasch models, the many-facet model, was applied to the analysis of these ratings. The analysis provided linear measures of the painfulness for each intensity of stimulation, of the pain perception of each subject and of the painfulness of each successive block. All these measures are located on a single pain perception continuum. Advantages and disadvantages of this methodology will be discussed in terms of subsequent possible mathematical analyses, statistical tests and implications for experimental and clinical investigations.