Bone Abstracts

Aim: The aim of the study was to evaluate the usefulness of the fracture risk evaluated through the FRAX® model based only on the clinical risk factors as a screening tool for identify the target population for treatment in osteoporosis.

Materials and methods: Two hundred and seventy-six postmenopausal women treatment naive referred to two different endocrinology departments for osteoporosis between 2009 and 2011 were evaluated. The FRAX® model for Romanian population was used to calculate the major osteoporotic risk fracture and hip risk fracture either using only clinical risk fracture (without BMD- abbreviated clinical FRAX®) or clinical risk fracture and femoral neck BMD (abbreviated FRAX®). We calculate the negative predictive value of clinical FRAX® outcome comparing to FRAX results to evaluate whether clinical FRAX® evaluation would be a reliable screening tool to select patient for completing DXA evaluation and initiate treatment in osteoporosis.

Results: The mean value of the clinical FRAX® evaluation in the study group was 7%±4.7 for the major osteoporotic fracture and 2.4%±2.7 for the hip fracture, respective. When femoral neck BMD was included, the mean value was 8.2%±5 for major osteoporotic fracture and 2.4%±2.7 for the hip fracture risk. Using the same cut off value (20% for major osteoporotic fracture and 3% for hip fracture risk) we found a positive predictive value of clinical FRAX® evaluation against complete FRAX® of 33% for major osteoporotic fracture and 77.6% for the hip fracture. The negative predictive value was found to be 95.9% for the major osteoporotic fracture and 90.05% for the hip fracture.

Conclusion: Using FRAX® evaluation based exclusively on clinical risk factors might be appropriate as a screening to identify patients with significant risk for major osteoporotic fracture but not for hip fracture. On the other hand we should highlight that our results are to be applied only in a high risk population like our study group of patients referred to university facilities.