Spinal Biomechanics Lab, Baylor College of Medicine, Houston, Texas, USA Documenting the patterns and frequency of collapse in non-operatively managed spine fractures, using a motion analysis software. Retrospective analysis of prospective case series 105 patients with thoracic or lumbar fractures, were neurologically intact, and treated non-operatively for the ‘stable’ injury at our unit between June 2003 and May 2006. The mean age of the cohort was 46.9 yrs Serial radiographs (mean 4 radiographs/patient; range 2 – 9) were analysed using motion analysis software for collapse at the fracture site. We defined collapse as a reduction of anterior or posterior vertebral body height greater than 15% of the endplate AP width, or a change in the angle between the inferior and superior endplates > 5°. The changes were assessed on serial radiographs performed at a mean of 5.6 mo (95% CI 4.1 – 7.1 mo) after the initial injury. 11% showed anterior collapse, 7.6% had posterior collapse, 14% had collapse apparent as vertebral body wedging, and 17% had any form of collapse. ODI scores were obtained in 35 patients at the time of the last available radiograph. There were no significant differences in ODI scores that could be associated with the presence of any form of collapse (p > 0.8 for anterior collapse; and p = 0.18 for posterior collapse). This pilot study with the motion analysis software demonstrates that some fractures are more likely to collapse with time. We hope to carry this work forward by way of a prospective study with a control on other variables that are likely to affect the pattern and probability of post-fracture collapse, including age, bone density, vertebral level, activity level, fracture type.
