Magnetic resonance imaging effectiveness in adolescent idiopathic scoliosis

Objective To assess the effectiveness of preoperative magnetic resonance imaging (MRI) in adolescent idiopathic scoliosis (AIS) patients with unremarkable history and physical examination. Methods The imaging data of consecutive patients with presumed AIS treated with a posterior spinal fusion between 2010 and 2016 were reviewed. The presence of traditional risk factors, atypical curve patterns, and its association with relevant abnormalities on MRI were investigated. The number needed to diagnose (NND) and the number needed to misdiagnose (NNM) were calculated to measure MRI effectiveness. Results A total of 198 consecutive patients were identified and divided according to the presence of MRI findings. Both groups predominantly consisted of females, with a mean age of 15 years and right thoracic curvature. Neural axis abnormalities were detected in 25 patients, and the groups had a similar proportion of atypical findings, as curve magnitude, thoracic kyphosis, curve direction, and sex. The NND was 7.9 patients and NNM was 66 patients, meaning that the management was changed before the spine fusion in 12% of patients with neural axis abnormalities. None of the traditional risk factors could predict a higher incidence of neural axis abnormalities in asymptomatic AIS patients. Conclusion Traditional risk factors may not be predictive of patients with a higher risk of changes in MRI. Both NND and NNM are representations easily understood by clinicians. Using these indexes to define if a patient should be submitted for additional imaging tests may facilitate the decision of using MRI as a preoperative screening tool in AIS patients. Level of evidence Level II