We are responsible for the care of almost 3000 patients with Behcet disease and read with great interest the report by Yasui and colleagues on successful treatment of this disease with pentoxifylline [1].

Behcet disease usually runs a course of exacerbations and remissions that gradually abate with time. Exacerbations in the eye cause a sharp decline in visual acuity that almost always improves as the attack subsides. Thus, studies testing treatment for the disease should not allow each patient to be his or her own control (as Yasui and colleagues' study did). Such a method results in a "regression toward the mean" bias. Therefore, if visual acuity is first measured during an attack, as seems to have been done in Yasui and colleagues' study, it will always improve when the attack subsides by itself or with the help of an effective remedy. A much more useful measure to follow is a baseline visual acuity, which reflects the visual acuity when an attack is not occurring.

The same observation also holds true for the data on neutrophil functions. The initial values were compared with those obtained at the second week of treatment. In addition, the patient with the highest value of O₂-release had been receiving colchicine before beginning pentoxifylline treatment. As the authors stated, colchicine by itself potentiates O₂-release.

Behcet disease is a serious health problem in some parts of the world, and the literature is replete with unsubstantiated remedies. We wholeheartedly agree with Yasui and colleagues that a more sophisticated study is necessary before any conclusions can be reached about the efficacy of pentoxifylline for treating Behcet disease.