To the Editor.
—The "randomized trial" by Van et al1 of diaper type in a day-care center is fundamentally flawed because vital information on potential biases has been omitted. Thus, the missing information should be requested from the authors and provided to the JAMA readership.In the study, the authors used a randomized, cross-over design. Presumably, the design was chosen to yield a definitive study of diaper type in the day-care setting. Instead, they chose to treat diapering like taking a drug. However, a "gold-standard" design for clinical trials is subject to significant biases in the social setting of the day-care environment. First, the care givers were not blinded, raising the question of what was the diaper preference of each of the care givers prior to the study. The authors gave no information on this fundamental point and did not mention it as a potential bias. It is up