An independent inter- and intra-observer agreement assessment of the AOSpine upper cervical injury classification system

Spine J. 2023 May;23(5):754-759. doi: 10.1016/j.spinee.2022.11.005. Epub 2022 Nov 15.

Abstract

Background context: The complex anatomy of the upper cervical spine resulted in numerous separate classification systems of upper cervical spine trauma. The AOSpine upper cervical classification system (UCCS) was recently described; however, an independent agreement assessment has not been performed.

Purpose: To perform an independent evaluation of the AOSpine UCCS.

Study design: Agreement study.

Patient sample: Eighty-four patients with upper cervical spine injuries.

Outcome measures: Inter-observer agreement; intra-observer agreement.

Methods: Complete imaging studies of 84 patients with upper cervical spine injuries, including all morphological types of injuries defined by the AOSpine UCCS were selected and classified by six evaluators (from three different countries). The 84 cases were presented to the same raters randomly after a 4-week interval for repeat evaluation. The Kappa coefficient (κ) was used to determine inter- and intra-observer agreement.

Results: The interobserver agreement was almost perfect when considering the fracture site (I, II or III), with κ=0.82 (0.78-0.83), but the agreement according to the site and type level was moderate, κ=0.57 (0.55-0.65). The intra-observer agreement was almost perfect considering the injury, with κ=0.83 (0.78-0.86), while according to site and type was substantial, κ=0.69 (0.67-0.71).

Conclusions: We observed only a moderate inter-observer agreement using this classification. We believe our results can be explained because this classification attempted to organize many different injury types into a single scheme.

Keywords: Agreement study; Classification system; Fracture classification; Spine; Upper cervical spine fracture.

MeSH terms

  • Humans
  • Lumbar Vertebrae* / injuries
  • Observer Variation
  • Reproducibility of Results
  • Spinal Injuries* / diagnostic imaging
  • Tomography, X-Ray Computed / methods