Brace Weaning Protocol Comparing Gradual and Immediate Weaning

The aims of this study are: 1) To compare the degree of Cobb angle and truncal balance maintenance between immediate removal and gradual brace weaning protocols for AIS patients who have completed brace treatment; 2) To determine factors for poor maintenance of Cobb angle and truncal balance after brace weaning; 3) To determine if the current standardized criteria for initiating brace weaning is appropriate.

The investigators' hypothesis is that an AIS patient requires gradual brace weaning to allow better adaptation to an unloaded environment without the brace thereby resulting with a better truncal balance. It is also hypothesized that the immediate removal of brace for a AIS patient is as effective in terminating bracing treatment when compared to gradual brace weaning over a 6-month period.

The investigators will recruit AIS patients managed at the Duchess of Kent Children's Hospital (DKCH) in Hong Kong. As there is no definition for "brace weaning", we adopt the protocol that our center has been using. Hence the investigators will be comparing two protocols: 1) brace wearing time shortened to night wearing for 6 more months before stopping bracing altogether or 2) immediate brace weaning on the day of patient recruitment. All patients will be recruited on the day of clinic consultation where the decision for brace weaning is made. Those with underarm (Boston) bracing are included in the study. All subjects are followed-up longitudinally up to 2 years after brace weaning has been issued. Hence this includes three clinic visits with one at 6 months to complete the gradual brace weaning protocol, and at 1 and 2 years to observe for any delayed changes to the overall curvature and balance including loss of truncal balance and Cobb angle deterioration. This does not deviate significantly from our usual care as all AIS subjects are usually followed-up at 6-monthly intervals and thus this duration of follow-up is strictly adherent to and standardized.

Eligible patients who have consented to take part will be randomly allocated to one of the two protocols, and group designation will be decided by a computer-generated randomized table. Restricted randomization through blocking will be used according to the equal proportion rule (1:1). The random allocation sequence will be managed independently, and the random coding is concealed to all research staff until all data analyses have been completed. The clinicians and research assistants will not be blinded to treatment allocation, as the participants are required to implement bracing protocols and schedule longitudinal follow-up.

Apart from the Cobb angle measurement, all subjects undergo weight, height and arm span measurements, health-related quality of life (HRQOL) and utility measurements, and whole spine standing posteroanterior (PA) and lateral radiographs at each clinic visit as per routine clinic protocol. The refined Scoliosis Research Society 22-item (SRS-22r) questionnaire and the EuroQoL 5-Dimension 5-Level (EQ5D) questionnaire will thus be provided at every visit.