Reliability of end, stable, neutral, first coronal reverse vertebrae identification in degenerative lumbar scoliosis: Intra- and interobserver consistency analysis

OBJECTIVE: To assess the intra- and interobserver reliability by observer training level used for selecting the end vertebra (EV), neutral vertebra (NV), stable vertebra (SV), and first coronal reverse vertebrae (FCRV) in degenerative lumbar scoliosis (DLS) patients. METHODS: Fifty consecutive upright long-cassette radiographs and CT examination of operative cases of DLS were evaluated by three surgeons at various levels of training. For each iteration, the observers attempted to identify the UEV, NV and SV from x-ray, and FCRV from the CT examination. Intra- and interobserver reliability was assessed by means of Cohen's Kappa correlation coefficient, and raw percentages of agreement were recorded. RESULTS: Intraobserver reliability was excellent for determining FCRV (K(a) = 0.761–0.837), fair to good for determining UEV (K(a) = 0.530–0.636), fair to good for determining SV (K(a) = 0.519–0.644), and fair to good for determining NV (K(a) = 0.504–0.734), respectively. Additionally, we also noted a trend towards better intraobserver reliability with increasing levels of experience. Interobserver reliability was poor between observers beyond chance for UEV, NV, SV (K(a) = 0.105–0.358), and good reliability for FCRV (K(a) = 0.581–0.624). All three observers agreed on the same level of the FCRV in 24 patients of the time, which presented less Coronal imbalance type C compared to the other 26 patients. CONCLUSION: Experience and training level of the observers are important factors affecting the accurate identification of these vertebrae in DLS, intraobserver reliability increases along with increasing levels of observer experience. FCRV is superior to UEV, NV, and SV in the accuracy of identification, Type C coronal malalignment could affect the accurate identification of FCRV.