To the Editor.
—Hartman and Landau1 compared "formal language therapy" with "supportive counseling" for aphasia. They did not test the efficacy of language treatment for aphasia. However, their comments and conclusions imply the authors believe they conducted an efficacy study, and, unfortunately, readers may infer that they did.Like most clinical trials, the Hartman and Landau effort contains some flaws. More importantly, their review of the literature contains major errors, and their conclusions are not supported by the results of the study they conducted.First, random assignment of "referred" patients to formal language therapy or supportive counseling does not yield "two randomly selected groups." The authors acknowledge that over 100 of the referred patients were rejected for a variety of reasons, and there were "at least as many very sick and very mildly affected patients who were not referred to us at all." It is appropriate to obtain samples from