To The Editor:
In the article entitled "Brace Wear Control of Curve Progression in Adolescent Idiopathic Scoliosis" (2010;92:1343-52), Katz et al. identified a correlation between greater brace wear time and lack of curve progression in patients with adolescent idiopathic scoliosis. However, one must question their conclusion that bracing controls scoliosis progression. A simple correlation does not prove a conclusion. Compliance may also reflect curve stability because the brace remains comfortable. Or, perhaps patients are willing to wear their brace as long as treatment appears to be successful. Those who observe their treatment to be failing may choose to stop that treatment and thus are labeled "noncompliant." A potential false assumption of treatment methodology is that treatment is benign, without intolerable side effects, and correlates with outcomes.
The true value of a treatment must be judged by the intent-to-treat for all patients who enter the study. All treated patients are then compared with a group of untreated patients. It should be noted that twenty-eight of the 100 patients who entered this bracing study progressed to surgery during the period of study. Goldberg et al.1, who did not utilize bracing as a treatment method, reported a 29% rate of surgery in premenarchal girls with scoliosis of 20° to 40° at the time of presentation. My colleagues and I reported a 22% rate of surgery in a long-term study of night-time bracing for scoliosis2.
While Katz et al. demonstrated that compliance correlates with curve stability, that observation does not necessarily indicate that bracing influences the outcome of scoliosis.