Bracing in adolescent idiopathic scoliosis, surrogate outcomes, and the number needed to treat.

BACKGROUND

In this manuscript we will present several evidence-based medicine concepts and tools that can be helpful to the clinician seeking answers to clinical questions. The clinical scenario used to demonstrate these concepts is one of the substantial current controversy in pediatric orthopaedics, that is the efficacy of bracing in adolescent idiopathic scoliosis. We hope to provide some important information about how to search and interpret the current literature on bracing but also to discuss the concepts related to "surrogate outcomes" and the "number needed to treat," which we believe are increasingly important in this era of evidence-based medicine.

METHODS

We performed a structured literature review of scoliosis bracing and also a separate analysis of the number needed to treat (NNT) for preventing surgery in adolescent idiopathic scoliosis.

RESULTS

Bracing for idiopathic scoliosis significantly reduces the rate of curve progression more than 6 degrees. However, the applicability of the 6-degree surrogate outcome compared with a more important outcome such as progression to surgery is doubtful. Bracing may decrease the risk of progression to surgery although the confidence intervals are large. The NNT for routine scoliosis bracing is about 9 patients for each surgery prevented. The NNT for patients highly compliant with bracing is about 4. We caution that these NNTs are derived from nonrandomized cohorts, and the true values from quality randomized controlled studies may be substantially different. There is no evidence for any particular brace over another although rigid bracing seems better than SpineCor bracing from 1 small randomized controlled study.

CONCLUSIONS

Systematic reviews support bracing's ability to prevent curve progression of 6 degrees but not for preventing surgery. Analysis of a patient cohort does support bracing's ability to prevent surgery with NNT of 9 for all patients and 4 for highly compliant patients.

LEVEL OF EVIDENCE

Systematic review-therapeutic level 2. Cohort analysis-therapeutic level 2.