Osteoporosis International vol:16 issue:1 pages:93-100
Identifying women with osteoporosis remains a clinical challenge, as it may not be feasible or cost-effective to recommend dual-energy X-ray absorptiometry (DXA) for all postmenopausal women. In this regard, quantitative ultrasound (QUS) has emerged as an attractive screening tool because of the (relatively) low cost and because QUS and DXA-assessed BMD appear to be equally predictive of future (hip) fracture risk. The objective of this study was to compare the ability of calcaneal QUS to identify osteoporosis with two alternative potential screening methods: digital X-ray radiogrammetry (DXR) and radiographic absorptiometry (RA). We enrolled a total of 221 postmenopausal community-dwelling Caucasian women aged 50-75 years. Bone mineral density (BMD) was measured at the lumbar spine and the total hip regions using DXA. Calcaneal ultrasound attenuation and velocity were assessed using QUS and metacarpal and phalangeal bone density were estimated by the use of DXR and RA, respectively. Receiver operating characteristic (ROC) curves were constructed by calculating the specificity and sensitivity of QUS, DXR, and RA at different cut-point values in discriminating osteoporosis, as defined by a T-score below -2.5 at the spine or hip using DXA, and the areas under the curves (AUCs) were computed. The sensitivity for identifying women with osteoporosis was 67.6% [95% confidence interval (CI), 50.2-82.0%] using QUS and was 76.9% (95% CI, 60.7-88.8%) and 82.9% (95% CI, 67.9-92.8%), respectively, using DXR and RA. The negative predictive value (NPV, the proportion of patients with a negative test who have no osteoporosis) was 90% for QUS, compared with an NPV of 94% for both DXR and RA. These data suggest that metacarpal DXR and phalangeal RA may be as effective as calcaneal QUS for targeting DXA testing in high-risk postmenopausal women.