L Rosenthall1. 1. Department of Nuclear Medicine, Montreal General Hospital, Que.
Abstract
OBJECTIVE: To determine the limitations on the improvements in the efficacy of screening with ultrasound densitometry of the calcaneus that can be achieved by adding direct-site mineral measurements of the lumbar spine and the femur when the ultrasound t-score is low. SUBJECTS AND METHODS: The author retrospectively analysed data from 2500 women for whom the results of both dual-energy x-ray absorptiometry of the lumbar spine and femur and ultrasonometry of the calcaneus were available. Various ultrasound t-score cut-off values (from 0 to -4.5) were tested to determine changes in sensitivity, specificity, false negatives, false positives and accuracy with and without adding direct-site measurements of the lumbar spine and the femur neck. RESULTS: For this analysis, the addition of direct-site measurement data increases the specificity to 100% for all t-score cut-off values without affecting the sensitivity, thereby improving the accuracy; however, the number of false negatives remains unknown. The number of false negatives decreases as the cut-off value is increased, but with higher cut-off values, greater numbers of subjects would have to be recalled for direct-site measurements of the lumbar spine and the femoral neck, which would entail extra costs. CONCLUSIONS: The accuracy of screening for low mineral content by calcaneal ultrasonometry is improved by recalling subjects for direct-site measurements if their ultrasound t-score falls below some arbitrary value. The cut-off value must be chosen to achieve a balance between the number of missed false negatives that are acceptable in the screening process and the number of subjects recalled for the direct-site study at additional cost. In remote areas, where dual-energy x-ray absorptiometry devices for performing the supplemental measurement are not usually available, the cost of transportation would be another factor.
OBJECTIVE: To determine the limitations on the improvements in the efficacy of screening with ultrasound densitometry of the calcaneus that can be achieved by adding direct-site mineral measurements of the lumbar spine and the femur when the ultrasound t-score is low. SUBJECTS AND METHODS: The author retrospectively analysed data from 2500 women for whom the results of both dual-energy x-ray absorptiometry of the lumbar spine and femur and ultrasonometry of the calcaneus were available. Various ultrasound t-score cut-off values (from 0 to -4.5) were tested to determine changes in sensitivity, specificity, false negatives, false positives and accuracy with and without adding direct-site measurements of the lumbar spine and the femur neck. RESULTS: For this analysis, the addition of direct-site measurement data increases the specificity to 100% for all t-score cut-off values without affecting the sensitivity, thereby improving the accuracy; however, the number of false negatives remains unknown. The number of false negatives decreases as the cut-off value is increased, but with higher cut-off values, greater numbers of subjects would have to be recalled for direct-site measurements of the lumbar spine and the femoral neck, which would entail extra costs. CONCLUSIONS: The accuracy of screening for low mineral content by calcaneal ultrasonometry is improved by recalling subjects for direct-site measurements if their ultrasound t-score falls below some arbitrary value. The cut-off value must be chosen to achieve a balance between the number of missed false negatives that are acceptable in the screening process and the number of subjects recalled for the direct-site study at additional cost. In remote areas, where dual-energy x-ray absorptiometry devices for performing the supplemental measurement are not usually available, the cost of transportation would be another factor.