RESUMEN
Nonpharmacological treatment of chronic pain in older people can be effective but attitudes and adherence to use of this treatment may differ by ethnicity. This study supports that a modified 14-item instrument based on the modified Health Belief Model-the arthritis-related health belief instrument (AHBI)-can be used across ethnically diverse older adults (i.e., European Americans, Hispanics, African Americans, and Afro-Caribbeans). Confirmatory factor analysis tested the factor structure of the AHBI to eliminate items inappropriate for this population. Structural equation modeling tested expected relationships among four latent variables-severity, susceptibility, barriers, and benefits-across the four ethnic groups. Findings suggest that the modified 14-item AHBI (eliminating two items from the original AHBI) adequately described the four latent factors pertaining to use of nonpharmacological pain therapy in this sample. All items registered substantial loadings (.41-.95) on the hypothesized factors, operating similarly across the four ethnic groups. The modified 14-item AHBI may be useful in (a) assessing how individual perceptions influence access to nonpharmacological pain therapy among ethnically diverse community-dwelling older adults, with the goal to develop and implement effective pain treatment for this population; and (b) measuring the likelihood of using nonpharmacological pain therapy by older adults. The modified 14-item AHBI can help health care providers to provide accurate pain assessment and examine domains that could affect use of nonpharmacological pain therapy by ethnically diverse older adults and guide practice with them by identifying barriers to use of such therapies and providing education to encourage their use.