TO THE EDITOR:
In "Adolescent Idiopathic Scoliosis: Treatment with the Wilmington Brace. A Comparison of Full-Time and Part-Time Use" (78-A: 1056—1062, July 1996), Allington and Bowen concluded that there is no significant difference between full-time and part-time use of the brace with regard to progression of the curve. While this may be true, I believe that their paper has several flaws that impair the scientific validity of their argument.
First, the authors only included compliant patients but never defined compliance or how they determined that the patients were compliant. Many of us who manage a large number of adolescent patients who have scoliosis believe that such patients tend to overstate the time spent in the brace when reporting to the treating physician. Allington and Bowen did not indicate how many patients were excluded for non-compliance. A treatment with which only 50 per cent of the patients are able to comply cannot be considered very effective.
Second, the authors did not state how the patients were assigned to the treatment groups; presumably, the patients were not randomized. Were the patients consecutive? Did the patients self-select their groups? Could patients who were non-compliant with full-time bracing be included in the group of patients who were managed with part-time bracing? While the demographics of the groups seem to be similar, randomization remains the scientific standard.
While I applaud the authors' efforts to increase compliance and to decrease any psychological side effects through the use of a part-time bracing program, I urge them to produce a more scientifically rigid study; the problem of documenting compliance continues to plague scoliosis research.
Malcolm L. Ecker, M.D.: 8815 Germantown Avenue, Suite 12, Philadelphia, Pennsylvania 19118
Dr. Allington and Dr. Bowen reply:
We thank Dr. Ecker for pointing out our omission of the criteria for determining compliance in the Materials and Methods section. This was an important error because compliance is an essential determinant of the effectiveness of the Wilmington brace.
We evaluated compliance in three ways: we questioned patients about compliance at the most recent follow-up evaluation, we determined the number of brace fabrications, and we reviewed the records to determine the orthopaedist's opinion regarding compliance. The patients were adults at the time of the most recent follow-up, and at least two years had passed since the completion of treatment with the Wilmington brace. They were questioned specifically about their compliance in a non-threatening way at the follow-up visit, and anyone who reported non-compliance was excluded from the study. We believe that the patients who were included answered truthfully and had no reason to lie.
The Wilmington brace is a total-contact thoracolumbosacral orthosis that is made of Orthoplast. The Orthoplast is only 3.2 millimeters thick, and the brace is not padded. When the brace is worn regularly, the Orthoplast absorbs perspiration and becomes brittle. The brace shows wear easily and lasts only about one year with intensive use. In fact, one of the major disadvantages of the Wilmington brace is its fragility. The patients who wore the brace full-time needed an average of three fabrications, and those who wore the brace part-time needed an average of two fabrications. We believe that compliance was demonstrated by the deterioration of the brace.
The charts were evaluated, and any patient who was reported to have been non-compliant was excluded from the study. The Wilmington brace offers the orthopaedist several opportunities to determine compliance. As the Wilmington brace is a total-contact orthosis, patients who wear the brace will almost always have skin changes. Thickening and hyperpigmentation of the skin may be noted over both anterior superior iliac spines and in the region of the brace molds that press against the skin over the apex of the curve. Also, a temporary deformity of the tenth, eleventh, and twelfth ribs develops in most patients after they wear the brace for several years. The brace is secured by circumferential nylon straps that squeeze the semiflexible Orthoplast against the chest. This results in a circumferential constriction of the lower ribs, which often can be seen on radiographs. We did not use deformity of the ribs as a criterion for inclusion because we do not know the reliability or sensitivity of this observation. Most likely, the review of the charts was the least accurate method that we used to evaluate compliance; however, all patients who were reported as non-compliant were excluded from the study.
Nanni J. Allington, M.D.: Centre Hospitalier Regional de la Citadelle, Boulevard du 12e du Ligne, 1, 4000 Liège, France
J. Richard Bowen, M.D.: Alfred I. duPont Hospital for Children, 1600 Rockland Road, P.O. Box 269, Wilmington, Delaware 19899