Brace wearing time is the strongest predictor of final results: A regression model in 1457 high risk consecutive adolescents with idiopathic scoliosis

Abstract

Current GUIDELINES propose PCA, but there are no large studies to check final results, and predict which patients will respond better. A personalised conservative approach (PCA) to adolescents with idiopathic scoliosis (AIS) is based on different treatment protocols according to risk groups (11–20, 21–30, 31–40, 41–45). The aim of the present research is to develop a model to predict end results of a PCA in adolescents with idiopathic scoliosis (AIS). Design: retrospective observational study nested into a prospective database. Inclusion criteria: AIS, 11–45°, Risser 0–2, age 10–16, first consultation, no previous bracing, at end of observation (Risser 3, medical prescription). Treatments followed a personalised conservative approach (PCA) following the step-by-step theory (Negrini, 2018): intensity increases with estimated risk factors. Outcomes: end Cobb angle < 50° and < 30° and no-progression. A backward selection regression modelling used to assess the effect of 7 covariates: age, BMI, ATR, Trunk Aesthetic Clinical Evaluation (TRACE) score, Risser and Cobb angle at baseline; referred brace wear (RBW) and risk groups according to which different PCA. In total, 1457 patients, 82.6% females, age 12.11 ± 1.05. End < 50° was predicted by BMI and RBW (0.21 and 0.10 probability respectively) while age, Cobb and ATR were statistically significant but weighting < 0.005. End < 30° is predicted by RBW (0.37), and Cobb (0.03), while age counts < 0.0005. No-progression was predicted by RBW (0.33); Cobb, TRACE and ATR counted < 0.02, and age < 0.0002. Considering the 4 risk groups, end < 30° and end < 50° probability decreases with the groups (R2 = 0.3 and 0.04, respectively). Time of brace wearing is the strongest predictor of final results whether a < 50, < 30 or stability outcomes are considered. Risk groups based on PCA are good predictors.