Abstract
PURPOSE: Variation in rib numbering has been noted in adolescent idiopathic scoliosis (AIS), but its effect on the reporting of fusion levels has not been studied. We hypothesized that vertebral numbering variations can lead to differing documentation of fusion levels. METHODS: We examined the radiographs of 161 surgical AIS patients and 179 control patients without scoliosis. For AIS patients, the operative report of fusion levels was compared to conventional vertebral labeling from the first thoracic level and proceeding caudal. We defined normal counts as 12 thoracic (rib-bearing) and five lumbar (non-rib-bearing) vertebrae. We compared our counts with data from 181 anatomic specimens. RESULTS: Among AIS patients, 22 (14 %) had an abnormal number of ribs and 29 (18 %) had either abnormal rib or lumbar count. In 12/29 (41 %) patients, the operative report differed from conventional labeling by one level, versus 3/132 (2 %) patients with normal numbering (p < 0.001). However, there were no cases seen of wrong fusion levels based on curve pattern. Among controls, 11 % had abnormal rib count (p = 0.41) compared to the rate in AIS. Anatomic specimen data did not differ in abnormal rib count (p = 1.0) or thoracolumbar pattern (p = 0.59). CONCLUSIONS: The rate of numerical variations in the thoracolumbar vertebrae of AIS patients is equivalent to that in the general population. When variations in rib count are present, differences in numbering levels can occur. In the treatment of scoliosis, no wrong fusion levels were noted. However, for both scoliosis patients and the general population, we suggest adherence to conventional labeling to enhance clarity.