Magnetic Resonance Imaging in Infantile Idiopathic Scoliosis: Is Universal Screening Necessary?

Study Design: Retrospective chart review. Objective: Determine if there is a subset of presumed infantile idiopathic scoliosis (IIS) patients who have a low incidence of neural axis abnormalities (NAAs) such that screening magnetic resonance imaging (MRI) may be delayed. Summary of Background Data: Individuals with presumed IIS have an increased incidence of NAA. Because of the increased incidence, screening MRI is recommended for all patients. We follow these guidelines at our institution. However, MRI screening in this age group is not without cost or risk. Methods: This is a retrospective study of 53 presumed IIS (onset ≤3 years) patients who had screening MRIs. Demographic and radiographic characteristics were collected. A binary regression using continuous and categorical variables was used to determine if a model could be created to accurately predict MRI necessity. A receiver operating characteristic (ROC) analysis was performed to determine if a threshold Cobb angle exists that is associated with an increased likelihood of NAA. Results: Of the 53 patients, 13 had NAA findings, resulting in a 24.5% incidence of NAAs. Significantly fewer abnormal MRIs were found in patients with Cobb angles <29.5° than those with Cobb angles >29.5° (13/33 [39%] vs. 0/20 [0%], p =.0008). Patients with Cobb angles >29.5° were 27 times more likely to have NAAs than those with angles <29.5° (odds ratio = 27.0 [95% CI = 1.5–486.0], p =.03). No other parameters have a predictive value for NAA (p >.05). Conclusion: This is the first study in IIS patients to identify a radiographic parameter that helps select out a subgroup for MRI screening. Additionally, we report an incidence of 24.5% NAAs in these patients, which is higher than previously reported.