To the Editor: Lovell et al. (Aug. 21 issue)1 report that adalimumab seems to be effective in the treatment of children with juvenile rheumatoid arthritis. However, interpretation of their results is difficult, given the medication-withdrawal design of the trial. During the open-label lead-in phase, patients who had a response to the drug were preselected and children who could not tolerate it were excluded. The double-blind phase was therefore enriched with patients who had a response to the drug, and the treatment effects observed may be larger than those seen in an unselected population. The evaluation of safety also raises some concerns. Since all the patients initially received adalim umab, how can we rule out a carryover effect that might have masked the true incidence of adverse events?2 The authors highlight the ethical issues associated with denying active treatment during a double-blind phase of placebo-controlled trials in pediatric populations. However, failure to appropriately estimate the potential benefit and side effects might be of even more concern. We believe that the assertion that “the openlabel lead-in approach is generalizable to clinical practice” must be considered with caution.