|
Objective:To evaluate the intraobserver and intreobserver reliability of the SLIC and Allen-ferguson classification systems of sub-axial cervical spine injury using clinical and imaging data that had not been premeasured. Methods:Preoperative imaging data(posteroanterior and lateral radiographs, CT,MR) and clinical data of 54 patients were evaluated by five orthopedic qualified surgeons. Assignment was repeated 6 weeks later with all the data presented in a different order. The results were determined by calculating the average percentage of intraobserver and interobserver agreement. Reliability and reproducibility was quantified by using kappa statistics and ICC. Results:Allen-ferguson classification demonstrated fair to excellent interobserver and intraobserver reliability. The mean interobserver reliability was 61.8%(mean Kappa coefficient,0.51),while intraobserver reliability was 68.7%(mean Kappa coefficient,0.60). SLIC systems have higher interobserver and intraobserver reliability in evaluating injury morphology and neurological function,while have poor interobserver and intraobserver reliability of DLC injury. Interobserver reliability assessed by mean percentage,kappa coefficients and ICC were 62.7.%,0.50 and 0.58±0.01 for injury morphology,55.8%,0.31 and 0.45±0.03 for DLC injury,69.7%,0.61,0.86±0.02 for neurological function,30.7%, 0.21 and 0.72±0.01 for SLIC total,74.2%,0.43,0.57±0.04 for therapy selection. Intraobserver reliability assessed by mean percentage,kappa coefficients and ICC were 71.8%,063 and 0.74±0.06 for injury morphology,68.9%,0.51 and 0.64±0.07 for DLC injury,79.6%,0.73,0.87±0.02 for neurological function,44.5%,0.39 and 0.81±0.04 for SLIC total,78.8%,0.59,0.75±0.09 for therapy selection. Conclusion:The mean to higher interobserver and intraobserver reliability present in evaluating Allen-ferguson classification,injury morphology, neurological function and therapy selection of SLIC systems. Allen-ferguson classification is not comprehensive. The main reasons for influence of the evaluation of DLC injury due to the uncertain judgement of ligament injury by using MRI.
|