Validation of the brief pain inventory for chronic nonmalignant pain

Abstract

The Brief Pain Inventory (BPI; Cleeland and associates) has been used primarily to assess patients with cancer-related pain. Although it has been validated in many languages and is widely used, there has not yet been research published to validate its use for patients with chronic nonmalignant pain as the primary presenting problem. This study was designed to fill this gap by examining the psychometric properties of the BPI in 440 patients with chronic intractable pain referred to a chronic pain clinic at a metropolitan tertiary-care teaching hospital. Results indicated acceptable internal consistency (Cronbach α coefficients were .85 for the intensity items and .88 for the interference items). A factor analysis resulted in 2 distinct and independent factors, supporting the validity of the 2-factor structure of the BPI. Zero-order correlations indicated that the association with a measure of disability (the Roland-Morris Disability Questionnaire [RMDQ]) was significantly higher for BPI interference (r = 0.57) than for BPI intensity (r = 0.40, t = 5.71, P < .01) and that the correlation with BPI interference was not more than 0.80, supporting the conclusion that these scales assess related, but also distinct, dimensions. Finally, the finding that both BPI scales showed statistically significant improvement with treatment confirms the responsivity of BPI in detecting and reflecting improvement in pain over time. Perspective This paper validated the psychometric properties of a pain Assessment instrument (The Brief Pain Inventory) originally developed to assess cancer pain and extended its use for the chronic nonmalignant pain population. This provides an important and widely used diagnostic tool for the clinician treating chronic pain. © 2004 by the American Pain Society.