The article by Bisgaard et al1in the December 2001 issue of CHEST provides some useful information concerning the cost effectiveness of inhaled corticosteroids in very young children with persistent asthma. However, I have concerns with the ethics of subjecting young children with persistent asthma to a placebo controller therapy. This study was an offshoot of a multicenter trial, published in 1999, that prospectively evaluated the efficacy of two different doses of fluticasone compared to placebo in young children with asthma.2 A pharmaceutical company sponsored the study, and one of the authors of the December 2001 paper was an employee of that company. In the initial prospective study, from which the cost-effectiveness data were obtained, 237 children, 1 to 3 years old, were randomized to receive fluticasone, 100 μg/d or 200 μg/d, or placebo for 12 weeks. Based on the inclusion criteria (having asthma symptoms or requiring albuterol on at least 7 days of the 14-day run-in period), these children had moderate persistent asthma. Patients could remain in the study if they had one exacerbation requiring prednisone but were withdrawn if they had two exacerbations. Cromolyn was the only controller medication that was allowed other than study drug, and it was unlikely that many patients were taking cromolyn. Thus, most of the 81 very young children randomized to the placebo group did not receive any controller therapy for 3 months. Is this ethical? It is well established that patients of all ages with moderate persistent asthma benefit from controller therapy. The study would have provided just as much useful information if it had dropped the placebo arm and simply evaluated the two doses of fluticasone. The study then would have required fewer patients and would have avoided the ethical issue of using a placebo.