TY - JOUR
T1 - Opportunistic CT screening of osteoporosis on thoracic and lumbar spine
T2 - a meta-analysis
AU - Zhu, Yeqing
AU - Triphuridet, Natthaya
AU - Yip, Rowena
AU - Becker, Betsy J.
AU - Wang, Yong
AU - Yankelevitz, David F.
AU - Henschke, Claudia I.
N1 - Publisher Copyright:
© 2021 Elsevier Inc.
PY - 2021/12
Y1 - 2021/12
N2 - Background: Osteoporotic fractures are a major contributor to late life morbidity and mortality, and impose a substantial societal cost, yet osteoporosis remains substantially underdiagnosed and undertreated. The purpose of this meta-analysis was to assess the pooled diagnostic sensitivity and specificity of computed tomography (CT) images for diagnosing osteoporosis in patients who meet WHO dual X-ray absorptiometry (DXA) osteoporosis criteria using specific Hounsfield unit (HU) values as a threshold. Methods: Systematic literature searches in PubMed, Embase, Web of Science and Google Scholar were performed from the earliest available date through 1 July 2018, restricted to publications in English. Participants in all studies underwent CT scans that included the lumbar and/or thoracic spine for different indications and HU measurements were used to identify osteoporosis. DXA scans served as the reference standard. Results: Ten eligible studies were identified. The mean area under the hierarchical summary receiver operating characteristic (ROC) curve for diagnosis osteoporosis was 0.84 (95% CI: 0.81, 0.87). The pooled diagnostic sensitivity and specificity of CT images to identify osteoporosis were 0.83 (95% CI: 0.73, 0.90) and 0.74 (95% CI: 0.69, 0.79). The positive likelihood ratio (LR+), negative likelihood ratio (LR-), and diagnostic odds ratio were 3.4 (95% CI: 2.7, 4.5), 0.21 (95% CI: 0.12, 0.36), and 16.4 (95% CI: 7.8, 34.3), respectively. The bias-adjusted sensitivity and specificity of CT were 0.73 and 0.71. Meta-regression demonstrated that country of study, DXA criteria and scanner manufacturer were significant factors associated with the sensitivity of CT in detecting osteoporosis while scanner manufacturer was the only factor associated with specificity of CT. Conclusions: This meta-analysis showed reasonable pooled sensitivity and specificity for using threshold values measured on CT scans to identify osteoporosis opportunistically.
AB - Background: Osteoporotic fractures are a major contributor to late life morbidity and mortality, and impose a substantial societal cost, yet osteoporosis remains substantially underdiagnosed and undertreated. The purpose of this meta-analysis was to assess the pooled diagnostic sensitivity and specificity of computed tomography (CT) images for diagnosing osteoporosis in patients who meet WHO dual X-ray absorptiometry (DXA) osteoporosis criteria using specific Hounsfield unit (HU) values as a threshold. Methods: Systematic literature searches in PubMed, Embase, Web of Science and Google Scholar were performed from the earliest available date through 1 July 2018, restricted to publications in English. Participants in all studies underwent CT scans that included the lumbar and/or thoracic spine for different indications and HU measurements were used to identify osteoporosis. DXA scans served as the reference standard. Results: Ten eligible studies were identified. The mean area under the hierarchical summary receiver operating characteristic (ROC) curve for diagnosis osteoporosis was 0.84 (95% CI: 0.81, 0.87). The pooled diagnostic sensitivity and specificity of CT images to identify osteoporosis were 0.83 (95% CI: 0.73, 0.90) and 0.74 (95% CI: 0.69, 0.79). The positive likelihood ratio (LR+), negative likelihood ratio (LR-), and diagnostic odds ratio were 3.4 (95% CI: 2.7, 4.5), 0.21 (95% CI: 0.12, 0.36), and 16.4 (95% CI: 7.8, 34.3), respectively. The bias-adjusted sensitivity and specificity of CT were 0.73 and 0.71. Meta-regression demonstrated that country of study, DXA criteria and scanner manufacturer were significant factors associated with the sensitivity of CT in detecting osteoporosis while scanner manufacturer was the only factor associated with specificity of CT. Conclusions: This meta-analysis showed reasonable pooled sensitivity and specificity for using threshold values measured on CT scans to identify osteoporosis opportunistically.
KW - Bone density
KW - CT attenuation
KW - Meta-analysis
KW - Opportunistic screening
KW - Osteoporosis
UR - http://www.scopus.com/inward/record.url?scp=85114712974&partnerID=8YFLogxK
U2 - 10.1016/j.clinimag.2021.08.005
DO - 10.1016/j.clinimag.2021.08.005
M3 - Article
C2 - 34530357
AN - SCOPUS:85114712974
SN - 0899-7071
VL - 80
SP - 382
EP - 390
JO - Clinical Imaging
JF - Clinical Imaging
ER -