Opportunistic screening of osteoporosis by CT scan compared to DXA: A systematic review and meta-analysis

The efficacy of opportunistic osteoporosis screening with computed tomography (CT) scans obtained for other indications has not yet been implemented by the current guidelines. We aimed to compile available evidence on the efficacy of osteoporosis screening with CT scans obtained for other indications compared with dual X-ray absorptiometry (DXA). Studies comparing the diagnostic performance of the CT scan with the DXA published before 2023 were retrieved. We conducted a bias assessment using the Newcastle-Ottawa Scale for cross-sectional studies. Correlation coefficients (CC), area under the curve (AUC), sensitivity, and specificity of the CT scans compared with the DXA were meta-analyzed with random effects modeling. 41 studies fulfilled the inclusion/exclusion criteria. The included studies reported weak to very strong CC (0.35 to 0.95) and low to high accuracy for opportunistic osteoporosis screening with CT scans. The meta-analysis showed a moderate pooled CC of 0.59 (95 % CI: 0.53-0.64, P-value<0.001), and a relatively high AUC of 0.81 (95 % CI: 0.78-0.84, P-value<0.001). Subgroup analysis based on age and menopausal status did not show significant between-group differences. Significantly higher accuracy measures were estimated for CT scans of the proximal femur compared to other anatomic regions (CC: 0.70, 95 % CI: 0.57-0.82; AUC: 0.79, 95 % CI: 0.72-0.87), North American cases (CC: 0.66, 95 % CI: 0.52-0.80; AUC: 0.82, 95 % CI: 0.82-0.83), and populations with a higher percentage of women (CC: 0.60, 95 % CI: 0.52-0.69; AUC: 0.86, 95 % CI: 0.83-0.89). We observed a moderate performance of opportunistic osteoporosis screening with CT scans obtained for other indications.