Abstract
Although the Journal's policy on the publication of negative results (Aug. 17 issue)1 is laudable, I cannot agree with Dr. Angell's assessment that the chosen example of a negative study2 was "carefully designed and analyzed." At least three flaws in the study of Mills et al. on the periconceptional use of vitamins and neural-tube defects2 are serious enough to render its results and conclusions untenable.