PSY67 Psychometric Validation of the Pain Impact Questionnaire Revised With a 1-Week Recall

Abstract

The Pain Impact Questionnaire Revised (PIQ-R) is a 6-item patient-reported measure developed with Item Response Theory and normed against a general US population sample (GPS). The PIQ-R yields the Pain Impact Score (PIS), and has been shown to have strong psychometric properties. The traditional standard PIQ-R uses a 4-week recall; this analysis tested the PIQ-R with a 1-week recall period. The shorter recall is useful for monitoring interventions with a relatively rapid impact on pain. Identical psychometric tests were run on the 1-week and 4-week PIQ-R using a representative GPS. Internal consistency was measured with Chronbach's alpha. Construct validity was tested first by correlating PIS with a self-reported 0-100 general health rating scale, then by calculating mean PIS by reported impact of pain condition and other conditions. Discriminant validity was demonstrated by correlating PIS with the Physical (PCS) and Mental (MCS) Component Summaries of the SF-12v2 Health Survey. Finally, the unidimensionality of the PIS was explored through factor analysis. Chronbach's alpha was .95 for 4-week recall and .96 for 1-week. Correlations between PIS and general health ratings were strong, at .60 for 4-week recall and .61 for 1-week. The mean PIS among people with back pain rose monotonically with the reported impact of the condition, for both 1-week and 4-week recall. Conversely, there was no such rise in PIS by impact of seasonal allergies, which do not generally cause pain. Among people reporting no medical conditions, mean PIS in both versions was well below the GPS mean. Both versions of the PIS correlated more strongly with PCS than with MCS, and exhibited a highly unidimensional structure. In all psychometric tests, the 1-week version of the PIQ-R showed nearly identical results to the 4-week version. Both versions show robust internal consistency, construct validity, discriminant validity and unidimensionality.