Psychometric validation of the Brief Pain Inventory-Short Form in patients with systemic lupus erythematosus in the United States.

Abstract

This study evaluated the Brief Pain Inventory-Short Form (BPI-SF) in patients with moderate-to-severe systemic lupus erythematosus (SLE). Patients ≥18 years old who self-reported a physician diagnosis of SLE (confirmed by medical record review) and active SLE (Systemic Lupus Activity Questionnaire (SLAQ) score of ≥11) were included. The BPI-SF and Short Form Health Survey version 2 (SF-36v2) were administered electronically at baseline, week 2 and week 12. Psychometric properties of the BPI-SF were evaluated. Cronbach alphas were >0.9 for all BPI-SF items, domains and total score. Test-retest reliability correlations for responses between baseline and week 2 of the BPI-SF had intraclass correlation coefficients (ICCs) ≥0.5. The BPI-SF domains and total score were moderately positively correlated to the SLAQ score (r ≥ 0.4), but negatively correlated to the SF-36v2 bodily pain domain (r ≤ -0.6). The BPI-SF domains and total score were moderately negatively correlated to the SF-36v2 physical functioning domain and physical component summary (r ≤ -0.4), with low correlations between the BPI-SF severity domain and SF-36v2 mental component summary (r = -0.16). Assessment of pain, as measured by the BPI-SF, demonstrated validity and reliability in a sample of patients with moderate-to-severe SLE.