Diagnostic efficacy of a novel bedside dual-energy X-ray absorptiometry device for the diagnosis of osteoporosis: a prospective cohort study

Abstract

Osteoporosis is a prevalent systemic skeletal disorder with reduced bone mineral density (BMD) and elevated fracture risk, constituting a major global public-health burden. Conventional axial dual-energy X-ray absorptiometry (DXA), the WHO-endorsed diagnostic gold-standard, requires dedicated radiology facilities, limiting its use for hospitalized, bedridden and elderly patients. This prospective cohort study (Xuzhou Medical University Affiliated Hospital, March-June 2026) assessed diagnostic performance of a novel portable bedside DXA for osteoporosis (WHO criterion: T-score ≤ −2.5). Outpatient peripheral and axial DXA served as reference comparators; peripheral DXA was not treated as diagnostic gold-standard, as WHO criteria rely primarily on central axial DXA. In total, 140 patients (mean age 68.2 ± 10.5 years, 91.4% female) completed four DXA scans within 72 h. An independent composite reference was built from three conventional DXA modalities, strictly excluding bedside DXA to avoid incorporation bias. Each conventional DXA modality was assessed against a leave-one-out independent reference standard that excludes the index modality itself. Pairwise head-to-head analyses between modalities were interpreted as comparative agreement between measurement tools rather than absolute diagnostic accuracy. Supplementary sensitivity analysis used a central-DXA-only WHO reference. Diagnostic metrics included AUC, accuracy, sensitivity, specificity, predictive values and Cohen’s Kappa. Against the WHO reference, bedside DXA yielded AUC = 0.733, accuracy = 64.29%, sensitivity = 86.11%, specificity = 41.18%, PPV = 60.78%, NPV = 73.68%, with fair inter-method agreement (Kappa = 0.276). When each conventional DXA modality was evaluated against its respective leave-one-out reference standard, diagnostic performance was substantially reduced: accuracy was 62.86% for outpatient peripheral DXA, 58.57% for lumbar-spine axial DXA, and 42.14% for femoral-neck axial DXA. Sensitivity analysis against central-DXA-only WHO reference yielded attenuated performance estimates. Bedside-DXA T-scores showed significant positive correlations with T-scores from all conventional modalities (all P<0.001). This novel bedside DXA possesses reliable diagnostic performance for osteoporosis when evaluated against an independent composite reference standard. As a practical point-of-care tool, it is promising for screening immobile inpatients, though peripheral-site measurements cannot fully replace WHO-standard central axial DXA assessment.