The Diagnostic performance of quantitative ultrasound calcaneus measurement in case finding for osteoporosis in Thai postmenopausal women

Abstract

Objectives: To assess the diagnostic performance of Quantitative ultrasound (QUS) calcaneus measurement in case finding for osteoporosis in thai postmenopausal women using Dual energy X-ray absorptiometry (DEXA) as gold standard Study design: A diagnostic test (Cross-sectional descriptive study). Setting: Nuclear Medicine division, Phramongkutklao Hospital. Research methodology: Determine the accuracy and reliability of the QUS calcaneous. 300 consecutive new postmenopausal women, who lost normal menstruation period at least 1 year, come for bone densitometry to evaluate possible osteoporosis at outpatient clinics of Nuclear Medicine Division, Phramongkutklao Hospital. Calcaneus bone density measurement was performed with an Achilles express ultrasound device (Lunar) and left femoral neck bone density was measured with Hologic QDR-4500 scanner (Hologic, Waltham MA). BMD were interpreted as osteoporosis or normal by WHO criteria for determining performance of QUS. Optimal cut-off values were determined by use ROC curve for diagnosis osteoporosis patients. Result: The number of osteoporosis in this study wa 107(35.67%), sensitivity was 39.25% specificity was 91.71%, positive predictive value was 72.41% and negative predictive was 73.14% when use Tscore of WHO criteria as a reference. The positive likelihood ratio in this study was 4.73 and negative likelihood ratio was 0.66. When used ROC curve determine the optimal cut-off values by Siffness index of the QUS and categorized age (<65or>_65) before did the test, sensitivity was 77.6% and specificity was 59.6% Conclusion: Bone mass density measurement for predicting osteoporosis by use QUS had a very low sensitivity, but had a high specificity when use DXA as a gold standard and use WHO T-score as a reference. When use Stiffness index of QUS find optimal cut-off values for identifying osteoporosis in postmenopausal women at ROC curve, sensitivity would be better, but not good enough for replace the standard tool (DXA).