Composite Pain Index: Reliability, Validity, and Sensitivity of a Patient-Reported Outcome for Research.

Abstract

A single score that represents the multidimensionality of pain would be an innovation for patient-reported outcomes. Our aim was to determine the reliability, validity, and sensitivity of the Composite Pain Index (CPI). Methodological analysis of data from a randomized controlled, pretest/post-test education-based intervention study. The study was conducted in outpatient oncology clinics. The 176 subjects had pain, were 52 ± 12.5 years on average, 63% were female, and 46% had stage IV cancers. We generated the CPI from pain location, intensity, quality, and pattern scores measured with an electronic version of Melzack's McGill Pain Questionnaire. The internal consistency values for the individual scores comprising the CPI were adequate (0.71 baseline, 0.69 post-test). Principal components analysis extracted one factor with an eigenvalue of 2.17 with explained variance of 54% at baseline and replicated the one factor with an eigenvalue of 2.11 at post-test. The factor loadings for location, intensity, quality, and pattern were 0.65, 0.71, 0.85, and 0.71, respectively (baseline), and 0.59, 0.81, 0.84, and 0.63, respectively (post-test). The CPI was sensitive to an education intervention effect. Findings support the CPI as a score that integrates the multidimensional pain experience in people with cancer. It could be used as a patient-reported outcome measure to quantify the complexity of pain in clinical research and population studies of cancer pain and studied for relevance in other pain populations.