Are High Risk Patients of Femoral Neck Fractures Predictable from the Severity of Osteoporosis?

Abstract

Proximal femur fractures (PFFs) are a major cause of disability in the elderly. Some reports say that osteoporosis is a factor contributing to PFFs and values of bonemineral density (BMD) can be used to predict PFFs. In this study we investigated thepossibility of using BMD monitoring to predict the risk of PFFs.This study compares a PFF group with a nonfracture control group. The selection criteria for both groups were female patients, age 60 and over, on whom dual energy xray absorptiometry (DXA) was performed. The PFF group was composed of 52 patients who were treated between Feb. 1997 and Nov. 1998 and DXA was performed within two weeks postoperatively. The nonfracture control group was made up of 27 patients who were treated for total knee replacement between April 1998 and Dec. 1998 and DXA was performed within two weeks preoperatively.Histograms were made using BMD data from lumbar and proximal femor measuring sites. The histograms were used to produce receiver-operating characteristic (ROC) curves. After analysing the histograms and ROC curves, two ranges of interest in each DXA report, neck and Wards were selected as the best sites to use when comparing the PFF and control groups.The two groups were compared to determine cut-off levels for PFFs. For the assessment of fracture risk of the patients with osteoporosis, we would like to propose for neck, 0.600g/cm2 BMD in which sensitivity is 76% and specificity is 80%. And for Wards, 0.400g/cm2 BMD in which sensitivity is 71% and specificity is 81%.