Adolescent Idiopathic Scoliosis: The Effect of Brace Treatment on the Incidence of Surgery.

Adolescent Idiopathic Scoliosis: The Effect of Brace Treatment on the Incidence of Surgery Goldberg CJ, Moore DP, Fogarty EE, et al (Children's Research Centre and Orthopaedic Department, Our Lady's Hospital for Sick Children, Dublin, Ireland), Spine. 2001;26:42-47.

According to the authors, studies of nonsurgical treatment for adolescent idiopathic scoliosis (AIS) typically have not been randomized or controlled and have reached equivocal conclusions. In addition, the use of a randomized controlled design for this condition has raised ethical concerns about the denial of treatment, and the development of a true placebo brace has been elusive. The purpose of this study was to determine whether fewer patients ultimately underwent surgery at an institution with a bracing policy for AIS than at the authors' center where the practice since 1991 has been to observe patients until they reach maturity or until surgery was indicated. The policy at the authors' center recommends surgery on the basis of the Cobb angle (an angle that measures scoliosis using radiographs), likelihood of further progression, cosmesis, maturity, patient and family preference, and the probability of benefit from the surgery. The authors also sought evidence that bracing would have prevented surgery in a sufficient number of cases to justify a change of protocol at their center.

The authors reviewed the medical records of 153 patients with AIS (142 girls, 11 boys). The patients were at least 10 years of age at diagnosis, were at least 15 years of age at last review, had a minimum Cobb angle of 10 degrees at or subsequent to diagnosis, had no other contributing condition, and had onset of symptoms after December 1990. Required documentation included Cobb angle at diagnosis, curve pattern, and menarchial and surgical status.

The mean curve at diagnosis was 32.95 degrees (SD=16.28). Sixty-one percent of the patients had major thoracic curves, 31% had thoracolumbar curves, 6.25% had lumbar curves, and 1% had double major curves. Of the patients who had operations (n=43; 28.1%), 30.8% had thoracic curves, 43.3% had thoracolumbar curves, and 16% had lumbar curves.

Comparison of the surgical and nonsurgical groups showed the mean age at diagnosis was not statistically significantly different. The surgical group had significantly higher Cobb measurements at diagnosis, at surgery, or at last radiograph and were older at last review than the nonsurgical group. For all Cobb angles, when the surgery rates of untreated girls (ie, unbraced) were compared with their menarchial status at diagnosis, those who were premenarchial at diagnosis had a higher incidence of surgery.

The researchers compared surgery rates for patients with different curve sizes at diagnosis with the rates reported by the bracing center. The Z test showed no significant difference for curves less than 50 degrees or greater than 60 degrees. The surgical rate for unbraced curves between 50 and 59 degrees, however, was significantly increased. Comparison of surgical incidence at the authors' center (28.1%) with that of an institution with an aggressive bracing policy (22.4%) did not show a significant difference.

The authors acknowledged that patients in this study could not be distinguished by curve pattern, sex, or maturity. Nonetheless, they concluded that, in terms of surgical incidence, wearing a compliant brace may reduce the Cobb angle, but there was no demonstrated advantage of bracing.

Diane R Katz, PT Tucson, Ariz