Background: Identifying factors that predict successful bracing in patients with idiopathic scoliosis may help planning treatment.
Aim: To assess predictors for long-term curve progression and health-related quality of life after Boston brace treatment.
Design: Observational retrospective cohort study with analysis of brace treated patients followed from start until at least 10 years after treatment.
Setting: Patients recruited from the country's entire population consecutively treated at the National Hospital.
Population: 365 patients (339 girls/26 boys) with idiopathic scoliosis. Mean (SD) chronological age/bone age at start bracing was 13.2 (1.9)/12.6 (1.9) years. The primary major curve measured 33.2 (7.4°), and the major levels were thoracic (N.=248), thoracolumbar (N.=78) and lumbar (N.=39). Mean bracing time was 2.8 (1.5) years. Long-term follow-up was in average 23.3 (4.1) years after weaning with a mean major curve of 33.0 (13.1°). Successful treatment was defined as a stable primary curve with progression ≤5°, and secondary the SRS-22 questionnaire assessed quality of life.
Methods: We applied linear or logistic regression with backward elimination. Internal validation was assessed by bootstrapping. Twelve variables were included in the prediction models: age, bone age, scoliosis in close family, major curve size, level, shape, flexibility and in-brace redressement, compliance, curve magnitude after 1 year, treatment time and quality of life (SRS-22).
Results: 290 patients (79%) had rated good compliance using the brace >20 hours daily. Treatment failure was observed in 65 patients (18%), and 27 of them were operated. The best baseline predictors were age and brace redressement. During treatment compliance, major curve after 1 year, and treatment time were the best predictors, while thoracic major curve, curve size at start bracing and scoliosis in close family also contributed to the final model. The model's ability to predict quality of life was low.
Conclusions: The best predictors for a long-term success were good redressement and compliance, unchanged or reduced major curve after one year and short treatment time.
Clinical rehabilitation impact: Predictors at baseline and during early treatment can help identifying patients who benefit from bracing.