Access to bone mineral density testing in patients at risk for osteoporosis

Prior studies have documented suboptimal diagnosis and treatment for osteoporosis in many settings. Consistent predictors of suboptimal management include patient age, physician training, and physician gender. We assessed whether access to bone mineral density (BMD) testing was a predictor of osteoporosis management in an at-risk population of patients from New Jersey.

Based on health care claims data, we identified three groups of at-risk beneficiaries, including women >or=65 (n=8,283), men and women >or=45 with a fracture (n=740), and men and women >or=45 taking chronic oral glucocorticoids (n=616). As the outcome of interest, we determined whether beneficiaries had undergone a BMD test and/or filled a prescription for a medicine used for osteoporosis (alendronate, calcitonin, hormone therapy, etidronate, risedronate, raloxifene, teriparatide) during the period 1 September 2002-31 August 2004. We assessed the relationship between this outcome and access to BMD testing. Access was characterized using two different measures: (1) the estimated driving time between each beneficiary's residence and the nearest BMD testing center ("driving time") and (2) the number of persons >or=65 years of age per BMD testing machine ("BMD scanner ratio") for each of the 21 counties in New Jersey.

Of the 9,640 beneficiaries, we found that 3,104 (32%) had undergone a BMD test, 2,893 (30%) had filled a prescription for an osteoporosis medication, and 4,364 (45%) had one or both. Across the 21 counties of New Jersey, the percentage of at-risk patients who had a BMD test and/or medication for osteoporosis ranged from 38 to 52%. In models adjusted for patient factors and the clustering of patients in counties, driving time was not associated with patients being screened or treated for osteoporosis. The BMD scanner ratio was a weak predictor of osteoporosis management.

Among beneficiaries of one large health insurer in New Jersey, two different measures of access to BMD testing were not important predictors of receiving testing and/or medications for osteoporosis.