Abstract
Fracture risk evaluation of postmenopausal women is suboptimal, but most women undergo screening mammography. Digital X-radiogrammetry (DXR) determines bone mineral density (BMD) at the metacarpal shaft and can be performed on mammography equipment. This study examined correlations between DXR and dual-energy X-ray absorptiometry (DXA) in women undergoing mammography, to identify optimal DXR thresholds for triage to osteoporosis screening by central DXA. Postmenopausal women over age 50 years, recruited from Westmead Hospital's Breast Cancer Institute, underwent mammography, DXR and DXA. Agreements were determined using the area under the receiver operator characteristic (AUC ROC) curve and Lin's concordance correlation coefficient. Optimal DXR T-scores to exclude osteoporosis by DXA were determined using the Youden's method. Of 200 women aged 64 ± 7 years (mean ± standard deviation [SD]), 82% had been diagnosed with breast cancer and 37% reported prior fracture. DXA T-scores were ≤ -1 at the spine, hip or forearm in 77.5% and accorded with DXR T-scores in 77%. For DXR and DXA T-scores ≤ -2.5, the AUC ROC was 0.87 (95% confidence interval [CI], 0.81-0.94) at the 1/3 radius, and 0.74 (95% CI, 0.64-0.84) for hip or spine. DXR T-scores > -1.98 provided a negative predictive value of 94% (range, 88%, 98%) for osteoporosis by central DXA. In response to a questionnaire, radiography staff responded that DXR added 5 minutes to patient throughput with minimal workflow impact. In the mammography setting, triaging women with a screening DXR T-score < -1.98 for DXA evaluation would capture a significant proportion of at-risk women who may not otherwise be identified and improve current low rates of osteoporosis screening. © 2022 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.