RESUMO
BACKGROUND: In clinical practice lateral images of the chest are performed for various reasons. As these lateral chest X rays show the vertebrae of the thoracic and thoraco-lumbar region, we wondered if these X-rays can be used for evaluation of vertebral fractures instead of separate thoracic spine X-rays. METHODS: To evaluate the agreement and intraobserver reliability of the semi-quantitative method for vertebral fractures on the lateral chest X-ray (X-chest) in comparison to the lateral thoracic spine X-ray (X-Tspine), two observers scored vertebral fractures on X-Tspine and twice on X-chest, separately,blinded and in different time periods. Agreement and Cohens' kappa were calculated for a diagnosis of any fracture on patient level and on vertebral body level. The study was done in patients visiting an outpatient geriatric day clinic, with a high prevalence of vertebral fractures. RESULTS: 109 patients were included. The intraobserver agreement for X-chest versus X-Tspine was 95-98%for the two levels of fracturing, with a Cohen's kappa of 0.88-0.91. The intraobserver agreement and reliability of the re-test on the X-chest showed an agreement between 91 and 98% with a Cohen's kappa of 0.81-0.93. More vertebrae were visible on the X-chest, mean 10.2, SD 0.66 versus mean 9.8, SD 0.73 on the X-Tspine (p < 0.001). CONCLUSION: The results show good agreement and intraobserver reliability on the X-chest compared to the X-Tspine for visualizing vertebral fractures. The results of this study emphasizes that the routinely performed X-chest is reliable in order to diagnose vertebral fractures.