ABSTRACT
I n t e r p r e t a t i o n . Forty-five women aged 79.1 ± 7.1 years with hip fractures were used as the hip-fracture group and were compared with 40 healthy controls aged from 65 to 87 years. In addition, 47 young controls, aged 20-40 years, were used as the reference population to express some of the results as T-scores. QUS measurements were performed with the Hologic Sahara, GE-Lunar Achilles+, and Sunlight Omnisense devices. The adjusted odds ratio (OR adj.) for a 1SD reduction showed that the Omnisense SOS was associated with a significant increase in fracture risk (OR adj. 2.83) comparable to the Sahara BUA (OR adj. 2.42) and Achilles BUA (OR adj. 3.29). However, given the large overlap between the 95% intervals of each odds ratio, no significant difference was found between the devices. Similarly, a comparison of the areas under the receiver operator characteristics (ROC) curves did not show any significant differences between the parameters. Using the default parameter for each device, the percentages of subjects correctly classified were 70%, 67.5%, and 62.5% for the Sahara, Achilles, and Omnisense devices, respectively. The combination of radius and calcaneum improves the OR (3.62 to 4.74), however the large confidence intervals do not allow a claim of superiority for one device over another. The three QUS technologies tested with respect to hip fractures seem to show the same discriminatory ability. However, there are some differences in diagnostic thresholds. Finally, the combination of several sites using two different devices is not clinically valuable.
