Does ScoliScore provide more information than traditional clinical estimates of curve progression?

STUDY DESIGN Retrospective study comparing ScoliScore and clinical risk estimates. OBJECTIVE The purpose of this study was to compare risk stratification between ScoliScore and traditional clinical estimates to determine whether ScoliScore provides unique information. SUMMARY OF BACKGROUND DATA ScoliScore is a genetic prognostic test designed to evaluate the risk of curve progression in skeletally immature patients with adolescent idiopathic scoliosis with Cobb angles of 10° to 25°. Clinicians are currently trying to better understand the role this test may play in guiding clinical decision making because current standards of curve progression are largely based on radiographical markers, such as curve magnitude and bone age. METHODS Ninety-one patients who received ScoliScore testing at our center and met study inclusion criteria were identified. Patients were given a "clinical risk" level using their Risser sign and Cobb angle. Assigned clinical risk levels were compared with the ScoliScore risk levels reported by the manufacturer's scoring algorithm. RESULTS ScoliScore risk distribution in our population was 36% low risk, 55% intermediate risk, and 9% high risk. This compares with 2%, 51%, and 47%, respectively, for comparable clinical risk groupings. Only 25% of patients were in the same risk category for both systems. There were no significant correlations between ScoliScore and age, race, menarcheal status, Risser sign, or sex. There was a positive correlation between the Cobb angle and the ScoliScore (r = 0.581, P < 0.001). Cobb angle remained significant in the multivariate regression model (P < 0.001), and Cobb angle was found to account for 33.3% of ScoliScore's variance. CONCLUSION Only Cobb angle showed significant correlation with ScoliScore among the socioclinical variables studied. The risk distribution of the 2 risk estimation systems examined differed markedly: ScoliScore predicted nearly 16 times more low-risk patients and more than 5 times fewer high-risk patients. This demonstrates that ScoliScore provides unique information to traditional predictors of curve progression, advancing our understanding of the role of ScoliScore in the clinical setting.