Clinical outcomes measurement: Initial validation of the Brief Pain Inventory for painful diabetic peripheral neuropathy (m-BPI-DPN)

Abstract

Listen

Translate article

This study evaluated the psychometric properties of a modified Brief Pain Inventory for painful diabetic peripheral neuropathy (m-BPI-DPN). The m-BPI-DPN is an 11-item numeric scale that measures pain severity and pain interference in seven functional areas. Patients with painful DPN (N=255), referred through several outpatient settings, completed a survey of health status, healthcare utilization, medication use, work productivity and patient-reported outcomes (m-BPI-DPN, SF-12v2, HADS, MOS Sleep Scale, Euro-QOL). Construct, convergent and discriminant validity and internal consistency reliability of the m-BPI-DPN were evaluated. Mean m-BPI-DPN Pain Severity was 4.75 (SD= 2.56); mean Interference was 4.90 (SD=2.82). Consistent with published studies, principal components factor analysis with oblique rotation revealed two interpretable factors (eigenvalues >1.0): a severity scale comprising the four BPI Severity items and an interference scale comprising the seven Interference items. Cronbach's alpha was high for both scales (.94) and item-total correlations were above .74 for all scale items. Mean Severity score was highly correlated with Bodily Pain from the SF-12v2 (rs=.625, p<.001), the Pain/discomfort item in the Euro-QOL (rs=.581, p<.001), and a Likert-type measure of pain severity (rs=.741, p<.001). Mean pain Severity and Interference were significantly associated with emergency room, primary care, pharmacist and nursing utilization, lost work productivity and satisfaction with pain medication. Individual m-BPI-DPN Interference domains were moderately correlated (rs's >.5, p<.001) with analogous measures and the Sleep interference item had a high, significant association with the three primary MOS-Sleep scales (rs's=.66-.71, p<.001). Forcing a three factor solution revealed two further interpretable subscales of the Interference scale, consistent with Mood/Sleep and Activity. Item-scale correlations provided modest evidence of discriminant validity, although the two subscales were highly correlated. Results indicate the m-BPI-DPN is a promising instrument in the evaluation of painful DPN. Research was supported by Pfizer, Inc.