Keele Research Repository

STUDY DESIGN: This is a retrospective longitudinal review. OBJECTIVE: The purpose of this review was to identify predictors of developing clinical scoliosis and compare between traumatic and neurological aetiologies of SCI. SETTING: This study was conducted at the Midland Centre of SCI. METHOD: Case notes of all patients injured at an age up to 18 years and admitted between 1971 and 2013 were reviewed. RESULTS: Sixty-nine individuals were identified, of which seven were excluded: three with pre-existing scoliosis and four with spina bifida. The remaining 62 (44 males, 18 females) had a median age at injury of 17 years (inter quartile range 13-17). Of these, 51 (82%) had traumatic and 11 (18%) had neurological injury. Most (42/51; 82%) of the children who had a traumatic injury were older than 13 years. The risk of developing scoliosis was lower for older patients (RR 0.68 per year, 95% CI 0.52-0.83) or following a traumatic injury (RR 0.36, 95% CI 0.20-0.66). A multivariable analysis based on age and trauma showed that only older age decreased the risk. A robust Receiver Operator Curve analysis suggested 14.6 years as the optimal threshold to predict development of scoliosis within 10 years (Area Under the Curve; AUC 0.83 (95% CI 0.73-0.93), sensitivity 70% (95% CI 50-89%), specificity 89% (95% CI 74-100%). CONCLUSION: Our results suggest that age below 14.6 years was a predictor for scoliosis. Once adjustment is made for age, the incidence of scoliosis does not differ between traumatic and neurological aetiologies of paediatric SCI injury.