Simplified criteria for selecting patients for vertebral fracture assessment.

ABSTRACT

The 2013 Position Development Conference of the International Society for Clinical Densitometry (ISCD) has adopted simplified indications for vertebral fracture assessment (VFA) based on an analysis of the Study of Osteoporotic Fractures (SOF). This showed that a simpler regression model, which included only age, bone mineral density (BMD), and height loss, was able to differentiate women with vertebral fractures from those without vertebral fractures almost as well as more complex models. We aimed to verify these findings in 1228 women referred for BMD testing and determine if the 2013 ISCD indications for VFA would perform as well the 2007 indications. The simple and complex SOF-based models were similar in terms of sensitivity (88.4% vs 89.4%), specificity (44.4% vs 45.5%), positive (25.9% vs 26.5%) and negative (94.5% vs 95.1%) predictive values, and area under the receiver operating characteristics curve (AUROC) (0.664 vs 0.674). The 2013 and 2007 ISCD VFA indications did not differ significantly in terms of sensitivity (88.2% vs 91.3%), specificity (41.3% vs 37.5%), positive (25.3% vs 22.9%) and negative (93.9% vs 95.5%) predictive values, and AUROC (0.648 vs 0.644). Our study provides support for the use of the simplified 2013 ISCD VFA indications as a practical approach to VFA testing.