The prevalence of diffuse idiopathic skeletal hyperostosis in Japan - the first report of measurement by CT and review of the literature.

BACKGROUND

Diffuse idiopathic skeletal hyperostosis (DISH) is prone to be accompanied by a spinal column fracture which is resistant to conservative therapy. This major characteristic of DISH is not recognized adequately by physicians, because the disease's detailed pathological condition has not yet been investigated. Therefore, the purposes of this study were to investigate the prevalence of DISH using computed tomography (CT), and to validate the reliability of CT interpretation.

METHODS

Subjects were 558 patients (300 male and 258 female) who underwent both CT of chest to pelvis and x-ray of chest and abdomen from August 2011 to July 2012 at any department other than orthopedic surgery in our institution. The definition of DISH based on x-ray as well as CT was the presence of consecutive fused vertebral bodies according to Resnick's criteria. The prevalence of DISH based on both modalities was calculated in all subjects. For 107 subjects extracted at random, intra- (Cohen kappa) and inter-observer error (Fleiss kappa) were calculated and the levels of fused segments were investigated.

RESULTS

Ninety-eight of 558 subjects (17.6%) were diagnosed as DISH by x-ray, and 152 (27.2%) by CT. Among males, 70 of 300 subjects (23.3%) were diagnosed by x-ray, and 116 (38.7%) by CT. Among females, 28 of 258 subjects (10.9%) were diagnosed by x-ray and 36 (14.0%) by CT. The levels of fused segments were presented from thoracic spine to lumbar spine, especially the middle and lower thoracic spine. Cohen kappa of x-ray was 0.587, and that of CT was 0.825. Fleiss kappa of x-ray was 0.552, and that of CT was 0.643.

CONCLUSIONS

The prevalence of DISH based on CT was 27.1%, which was higher than that of x-ray. In addition, intra- and inter-observer error by review of CT was less than that of x-ray. CT evaluation would be a better method for precise understanding of the state of DISH.