Psychometric features of brief pain inventory for Parkinson's disease during medication states.

Purpose Patients with idiopathic Parkinson's disease (PD) suffer from different non-motor symptoms, including pain. The present study aimed to measure the psychometric properties of the Brief Pain Inventory (BPI) in patients with PD during ON- and OFF-states. Methods We recruited 460 patients with PD and 100 non-PD controls. The pain was assessed by the BPI, King's Parkinson's disease Pain Scale (KPPS), Neuropathic Pain Symptom Inventory (NPSI), Visual Analogue Scale-Pain (VAS-pain), and short-form McGill Pain Questionnaire-2 (SF-MPQ-2) in both medication states. Internal consistency and test-retest reliability was examined using Cronbach's alpha coefficient and intra-class correlation coefficient (ICC). Dimensionality and convergent validity of BPI were also investigated. Diagnostic accuracy and discriminative validity were determined by Receiver Operating Characteristics (ROC) curve analysis and Area Under the Curve (AUC). Results Cronbach's alpha was satisfactory (α = 0.91-0.97) in both states. The ICC values were 0.85-0.96 in ON- and OFF-state. Factor analysis revealed two factors. A high correlation was obtained between BPI subscales and other scales. AUC >0.91, sensitivity, and specificity> 0.77 were observed for discriminating different pain levels. Furthermore, appropriate diagnostic accuracy was found (AUC, sensitivity, and specificity >0.67) between non-PD control and PD patients. Conclusion The BPI has acceptable psychometric features as well as diagnostic accuracy for patients with PD.Implications for rehabilitationPain as a non-motor symptom in PD can affect daily and social activities.The BPI is used to assess pain severity and interference in activities.For better treatment, pain should be assessed in off-state like to on-state.BPI has satisfactory reliability and validity in different medication states in PD.