Comparison of TLICS & McAfee Classification in Thoracolumbar Injuries

ABSTRACT

STUDY DESIGN: A retrospective study. OBJECTIVES: We assessed the intraobserver and interobserver reliability of TLICS classification in the thoracolumbar injuries, which had been evaluated in our hospital. It was compared with that of the older, McAfee classification and discussed for clinical validation. SUMMARY OF LITERATURE REVIEW Among the numerous literatures regarding the thoracolumbar injury, there is no consensus on the most useful classification, and there is nothing comparing the McAfee classification with the TLICS classification. MATERIALS AND METHODS: Among the 230 patients that were treated with conservative care or operation from January 1, 2005 to January 1, 2010 in our hospital, 185 patients with initial CT and MRI images were assessed. Five orthopedic surgeons reviewed histories, plain film, CT and MRI of the 185 thoracolumbar injury cases, respectively. Each case was classified and scored according to the McAfee classification and the TLICS classification. The case assessment was recorded and the orthopedic surgeons repeated the assessment 1 month later. Intraobserver and interobserver reliability were assessed by statistical analysis. The actual management of each case was compared with the treatment recommended by TLICS classification to calculate the validity of the indexes. RESULTS: Intraobserver and interobserver reliability in TLICS were higher than those in the McAfee classification. Agreement of the TLICS classification for treatment recommendation was 81.7%, comparing with the actual management of previous McAfee classification. Validity indexes were satisfactory in therapeutic decision making, especially specificity. CONCLUSIONS: TLICS classification has a relative high K-value, when compared with that of the McAfee classification for intraobserver and interobserver reliability. Through clinical studies, including prospective observational analysis, TLICS classification can be applied and adjusted more adequately.