Is time since hip fracture influencing the discrimination between fractured and nonfractured subjects as assessed at the calcaneum by three technologically different quantitative ultrasound devices?

Because quantitative ultrasound (QUS) instruments from different manufacturers have significant technical differences, it is difficult to assess whether all of them can discriminate similarly between osteoporotic fractures and age-matched controls. Thus, to avoid any bias, reliable comparative assessment of the QUS devices should be carried out on the same population. Few studies have fulfilled this condition. Another source of variability in cross-sectional studies in which fractured and nonfractured subjects are compared is the time since osteoporotic fracture. Our study evaluated the ability of three calcaneal QUS devices to discriminate patients with osteoporotic hip fracture from control subjects, using the same population. In addition, a subset of patients was re-measured about 9 months after the hip replacement surgery to check how the time since fracture affects the discriminatory ability of the different QUS devices. Fifty postmenopausal hip-fractured patients and 46 postmenopausal age-matched controls were included in this study and measured on three QUS devices, as well as 50 young healthy controls to calculate the T-score. Odds ratio results showed that a decrease in UBIS(TM) BUA of I SD was associated with a significant increase in fracture risk (odds ratio adjusted = 2.30) comparable with Sahara broadband ultrasound attenuation (BUA) (OR adj. = 2.30), and Achilles BUA (OR adj. = 3.5). However, given the large overlap between the 95% intervals of each OR and for the areas under ROC curves, no significant difference was found between them. In the subset of 15 hip-fractured subjects, no significant differences were found between ultrasound parameters of the first visit and 9 months after except for the heel width (soft tissue variation). Odds ratio and areas under the curve (AUC) tend to increase from visit I to 2 for the BUA and decrease substantially for the SOS for all but the Lunar Achilles+. Nonsignificant correlation was found between the absolute difference of the ultrasound parameters measured at the two visits and the time since fracture, except for the Sahara SOS (r = 0.45; P < 0.04). In conclusion, no significant differences between QUS technologies were observed in their positive and significant ability to discriminate hip-fractured patient from controls. However, this statement is shadowed when taking into account the time since fracture which seems to negatively influence results obtained on dry versus wet QUS systems. As a result, it is advisable that such parameters would be taken into account when designing a study aimed to demonstrate the discriminatory ability of heel ultrasound between normal and hip-fractured patients.