Comparative analysis of the inter-observer reliability and intra-observer reproducibility between Lenke and PUMC classification systems for adolescent idiopathic scoliosis

Abstract

To compare the intra-observer reproducibility and inter-observer reliability regarding the results of Lenke's and Peking Union Medical College (PUMC) classification systems for adolescent idiopathic scoliosis (AIS). Five spine surgeons independently measured the X ray films of 62 AIS patients using both Lenke and PUMC classification systems. Three weeks later, the spine surgeons repeated the same classification process. No measurement trace was allowed to be left on the X ray films. SAS software was used to calculate the Kappa values and analyze the intra-observer reproducibility and inter-observer reliability The overall reliability and reproducibility rates of the Lenke classification system were 69.8% (Kappa value = 0.675) and 74.2% (Kappa value = 0.690) respectively, in which the curve characteristic parameter reliability and reproducibility rates were 86.5% (Kappa value = 0.808) and 87.4% (Kappa value 0.826) respectively, the lumbar correction reliability and reproducibility rates were 95.2% (Kappa value = 0.919) and 94.5% (Kappa value = 0.908) respectively, and the sagittal thoracic correction reliability and reproducibility rates were 85.2% (Kappa value = 0.734) and 89.0% (Kappa value = 0.805 respectively). The reliability and reproducibility rates of the PUMC classification system were 91.0% (Kappa value = 0.896) and 90.2% (Kappa value = 0.892) respectively. The reliability and reproducibility rates of the Lenke classification system were better than the previously reported results. However, the PUMC classification system expresses better reliability and reproducibility. The PUMC classification system is relatively easier with fewer curve types, hence leading to less divergence among the clinicians. Comparing to the classifications used in the past, both classification systems have the advantages of correctly categorizing AIS according to its individual traits, as well as accurately examining the X ray results, hence achieving significance in terms of surgical outcome and design.