LAUSR

BACKGROUND TLICS (Thoracolumbar Injury Classification System) and TLAOSIS (Thoracolumbar AOSpine Injury Score) are the scores preferred to classify and treat thoracolumbar fractures. Our study tries to evaluate the reliability of both as guidelines for treatment. METHODS Single center and retrospective case series of 458 patients. Clinical variables, radiology and treatment was analyzed. We classified fractures according to the AOSpine Thoracolumbar System and retrospectively applied both scales in two groups: surgical and conservative. A concordance analysis and statistical measures comparing both was performed. RESULTS 458 patients divided in two groups: 257 (56.1%) conservative and 201 (43.9%) surgical. The concordance analysis between both scales was 89.7% (CI95%: 86.5% - 92.3%) and Cohen's kappa coefficient was 0.68 (0.59 - 0.76). TLAOSIS had a higher tendency to classify patients in the gray zone (10.3% vs. 2.8%, p<0.001), while TLICS had a more conservative nature (85.2% vs 78.4%, p=0.01). In the surgical group the matching-decision ratio was 29.9% for TLICS and 42.8% for TLAOSIS but differences were found in TLICS being more conservative (70.1% vs 57.2%, p=0.01). In the conservative group the matching-decision ratio was 98.1% for both scales, being the main difference in the gray zone for TLAOSIS. CONCLUSION Both scales have a good concordance in general, with TLICS being more conservative overall. They had rather low coincidence when predicting surgery. Since TLAOSIS placed more patients in the "gray zone" we believe it might be slightly better for giving surgeons more "license" to decide a surgical approach on certain controversial types of fractures.