The role of diffuse idiopathic skeletal hyperostosis (DISH) in positive identification

The use of skeletal pathologies in establishing positive identifications via radiographic comparison is often avoided-and thus understudied-due to the dynamic nature of the skeleton in response to pathological conditions. Using an online survey, this study tests the accuracy of diffuse idiopathic skeletal hyperostosis (DISH), a relatively common vertebral pathology, in making positive identifications through radiographic comparison. Three digital radiographic images from 51 DISH-positive individuals were obtained from the Boston Medical Center: one image taken at a baseline date (Group A), one image taken within 2 years from baseline (Group B), and one image taken greater than 4.5 years from baseline (Group C). Survey participants were tasked with comparing between simulated lateral "antemortem" and "postmortem" images from living patients and identifying which pair represented the same individual at different time intervals. A total of 40 responses were recorded and analyzed by measuring accuracies, sensitivities, and specificities. Information about survey participants' field, degree, experience working with radiographs, and familiarity with DISH was also recorded. Series 1 compared Group A to Group B images and resulted in an accuracy of 87.3%, sensitivity of 46.9%, and specificity of 94.3%. Series 2 compared Group A to Group C images and resulted in an accuracy of 83.4%, sensitivity of 34.3%, and specificity of 95.8%. The results indicate that the progressive changes associated with DISH render the condition unreliable for making positive identifications but suggest that DISH could be used to winnow potential matches.