Vertebral Fractures: Which Radiological Criteria Are Better Associated With the Clinical Course of Osteoporosis?

Dear Dr Patlas, We thank Dr Wang for his interest in our paper and kind words. Readers may not know that, purely by chance, within months 3 groups separately published upon the diagnosis of osteoporotic vertebral fractures from plain radiographs. Their results were similar and have been reviewed elsewhere. Dr Wang and his colleagues at the Chinese University of Hong Kong was one of these groups; the other 2 were Drs L and E. Oei, and F. Rivadeneira et al at the Erasmus University, Rotterdam and the CaMos Consortium in Canada. All 3 compared morphometric (measurement-based) and morphologic (structural-based) diagnostic methods. In total, the studies included over 5000 patients, most from Canada. Although there is broad agreement between the findings of these groups, protocols and practices differed slightly. For example, when Jiang and her colleagues at the University of Sheffield in the United Kingdom first developed a structured approach to fracture diagnosis they observed a counter-intuitive difference in the segmental distributions of morphologic and morphometric deformities in the spines of osteoporotic patients. That observation was repeated by the Dutch and Canadian investigators but not by those in Hong Kong. It transpires that the Hong Kong group had intuitively concluded that morphometric abnormalities in the thoracic spine were not of osteoporotic provenance and tended to lead to over diagnosis. For this reason such ‘‘lesions’’ were not included in their data. Because of such differences in diagnosis and in the terminology used there appears to be a need for a nosology of vertebral deformities and confounders in this context so that there is a consistency in identifying and reporting findings. It may seem quaint to be earnestly discussing plain image radiography in this era in which radiological innovation is dominated by the power of sectional imaging. However, plain spinal radiography plays an important clinical role in future fracture-risk assessment as part of osteoporosis care. Dr Wang’s observations suggest there may be yet more to learn from these images, although the Japanese work has yet to be validated. Not least, as one reflects on the chequered history of osteoporotic vertebral fractures diagnosis and as artificial intelligence and machine learning may come to play an increasing role in radiology practice, it will be even more important to look critically at received wisdom such that human misunderstandings are not perpetuated in machine protocols.