A novel classification of osteotomized debridement based on the range of focus in treating active thoracolumbar tuberculosis: a multicenter study.

INTRODUCTION

Osteotomized debridement (OD) is increasingly used in the treatment of active thoracolumbar tuberculosis (TB). So far, no nomenclature has been established to describe the patterns of OD, and thus the surgical outcomes cannot be directly analyzed and compared among the patients treated with different extents of OD. The purpose of this study was to establish a reliable classification of OD for further study of spinal TB.

MATERIALS AND METHODS

This was a multicenter retrospective study. The proposed classification included 6 grades of OD based on sagittal range of vertebral body destruction: grade 0 involves single-level intervertebral disc and adjacent superficial endplates; grade 1 involves adjacent endplates and vertebral bodies, but no pedicle is involved; grade 2 involves adjacent endplates, vertebral bodies, and a lower or upper pedicle; grade 3 involves adjacent endplates, vertebral bodies, and both of lower and upper pedicles; grade 4 involves an entire vertebral body and an adjacent lower or upper pedicle; grade 5 involves two continuous entire vertebral bodies. Two hundred and five patients with active thoracolumbar TB who underwent OD surgery were included, and all ODs were classified. The reliability of this classification was evaluated twice by 10 readers, and Fleiss kappa coefficients were calculated.

RESULTS

In the 205 patients, 208 ODs were performed. Grade 2 OD was the commonest type (98/208, 47.1%), followed by grade 1 (50/208, 24.0%), grade 3 (26/208, 12.5%), grade 0 (20/208, 9.6%), grade 4 (8/208, 3.8%), and grade 5 (6/208, 2.9%). The average accuracy of the two readings was 86.2% and 90.1%, respectively. The intra-rater reliability for the classification was "almost perfect agreement" with a Fleiss kappa coefficient average of 0.92. The inter-rater reliability was "almost perfect agreement" with a coefficient average of 0.89 for two readings.

CONCLUSIONS

This classification proved to be intuitive and reliable. The graded OD provides a platform for preoperative evaluation and allows comparative analysis of clinical outcomes in different extents of OD.