Association between braced curve behavior by pubertal growth peak and bracing effectiveness in female idiopathic scoliosis: a longitudinal cohort study

BACKGROUND: Pre-pubertal idiopathic scoliosis (IS) is associated with high risk of bracing ineffectiveness. Integrated multidimensional maturity assessments are useful but complex to predict the high-risk occurrence of curve progression. This study is designed to provide a simple screening method for brace effectiveness by determining whether or not the braced curve behavior at growth spurt, being defined as variations in Cobb angle velocity (AV) at peak height velocity (PHV), can be a new factor predictive of brace outcome prescribed before PHV. METHODS: This is a retrospective study of a series of 35 IS girls with simplified skeletal maturity score no more than 3 at initiation of bracing treatment and followed up through the growth spurt until brace weaning or surgery. Serial Cobb angle and maturity indicators involving height velocity, Risser sign, triradiate cartilage, simplified skeletal maturity score and distal radius and ulna classification were assessed and patients were stratified into either a positive or negative category based on a positive or negative value of AV at PHV. Comparisons were made between the positive and negative AV groups, as well as the failed and successful bracing groups, using independent sample T test and crosstab analysis. Logistic regression analysis was used to identify the predictive factors of failed brace treatment. RESULTS: Brace treatment prescribed before PHV was found to have an overall failure rate of 57.1% and a surgical rate of 45.7%. Negative AV at PHV accounting for 54.3% of the recruited patients were associated with lower brace failure rate (36.8% vs. 81.2%, p = 0.016) and surgical rate (21.1% vs. 75.0%, p = 0.002). Patients in the failed bracing group showed higher ratio of thoracic curve (80.0% vs. 26.7%,p = 0.002) and higher AV at growth peak (2.3 ± 9.1 vs. -6.5 ± 11.4°/yrs., p = 0.016). The logistic regression analysis revealed that positive AV at PHV (OR = 9.268, 95% CI = 1.279–67.137, p = 0.028) and thoracic curve type (OR = 13.391, 95% CI = 2.006–89.412, p = 0.007) were strong predictive factors of ineffective brace treatment initiated before PHV. CONCLUSIONS: Sustained curve correction following bracing despite early onset and rapid pubertal growth was strongly predictive of effective brace control of scoliosis.