Reliability and reproducibility analysis of clinical and radiographic degenerative spondylolisthesis classification

English Keywords:Lumbar spondylolisthesis  Degeneration  Classification  Reliability  Reproducibility

English Abstract:

  【Abstract】 Objectives: To investigate the inter-observer reliability and intra-observer reproducibility of the clinical and radiographic degenerative spondylolisthesis(CARDS) classification system for degenerative spondylolisthesis and its clinical values. Methods: From January 2010 to December 2013, 49 patients with degenerative spondylolisthesis were treated in our department. Based on the CARDS classification system, the radiographs were assessed by five observers independently on two separate occasions with an interval of 1 month. The inter-observer reliability and intra-observer reproducibility of the system were quantified by using Kappa value and Spearman correlation coefficient. Results: The total 490 classifications were evaluated for two times. Fifteen percent(74 cases) were described as type A; thirty-four percent(167 cases), as type B; thirty-eight percent(187 cases), as type C and thirteen percent(62 cases), as type D. The inter-observer reliability assessed by the mean percentage and Kappa values at the first time was 77.6%-93.9%(89.6%), 0.764 respectively for disc space; 69.4%-91.8%(86.6%), 0.719 respectively for kyphosis; 67.3%-81.6%(77.5%), 0.596 respectively for translation; 65.3%-81.6%(74.1%), 0.641 respectively for final classification. The Spearman correlation coefficient was 0.564, 0.612, 0.576 and 0.589, respectively. The intra-observer reproducibility assessed by the mean percentage and Kappa values was 85.7%-98.0%(94.7%), 0.857 respectively for disc space; 89.8%-95.9%(92.1%), 0.842 respectively for kyphosis; 77.6%-91.8%(87.5%), 0.779 respectively for translation;75.5%-87.8%(82.4%), 0.804 respectively for final classification. The Spearman correlation coefficient was 0.714, 0.702, 0.689 and 0.645, respectively. Conclusions: The CARDS classification system for degenerative spondylolisthesis has good inter-observer reliability and excellent intra-observer reproducibility.