Infantile Idiopathic Scoliosis: Factors Affecting EDF Casting Success

Abstract

Study DesignIRB-approved retrospective single cohort study. ObjectivesTo review our ten-year history with EDF (Elongation Derotation Flexion) casting in patients with infantile idiopathic scoliosis (IIS) to better understand which factors predict successful outcomes. Summary of Background DataNumerous studies have demonstrated the efficacy of EDF casting in the treatment of progressive infantile idiopathic scoliosis. But none have reproduced the success of Mehta’s even with early intervention. MethodsPatients with IIS treated with EDF casting with a minimum 24-month follow-up were included. Radiographs and clinical records were reviewed. Age, sex, and curve type were documented. Precast, traction, in cast, in brace, and final Cobb angles were measured and recorded. Outcomes were defined by Cobb angle at final follow-up out of cast or brace. Patients were considered cured if the final Cobb angle was <10°, palliated at 10°–45°, and failed if they required surgical treatment. ResultsSixty-three patients with IIS were reviewed. Thirty-two were excluded for incomplete records or insufficient follow-up, leaving 31 patients. No patients progressed to surgical intervention during the study. Patients with a Cobb angle >10° in the final cast were 7.3 times more likely to fall into the palliative range at the most recent follow-up than if their Cobb angle was 10° or less even when adjusted for age. Earlier age at onset of casting (14.9 vs. 21.1 months) was not statistically significant (p=.073). Magnitude of initial curvature, flexibility, initial correction, sex, and curve type were also not found to be significant. ConclusionEDF casting is a valid treatment option for IIS with a high cure rate. Major Cobb angle at the end of casting is most predictive of outcome. Cobb angles >10° at the end of casting had a 7.3 times greater risk of falling into the palliated category versus Cobb angles less than or equal to 10° even when adjusted for age. Initial curve magnitude, curve flexibility, sex, and curve type were not predictive. Level of EvidenceLevel IV.