To the Editor.
—Dr Ross-Degnan and colleagues1 are to be commended for bringing to the attention of physicians the unfortunate results of drug company marketing on the prescribing practices of US physicians.However, the logic of Ross-Degnan et al weakens when they state that the risks of drugs substituted for a drug withdrawn because of unanticipated bad effects should be considered before withdrawing the drug. The drug would not have been released if the bad effects had been known beforehand. Therefore, patients who needed to be treated for conditions for which the drug was putatively indicated would have been given the substituted drugs as first choice. Ergo, no net effect.