To the Editor.—The admonitions of Adson1 in his recent presidential address on the misuse of quantitative data in reports on clinical research are sorrowfully ignored several pages on in the same issue of the Archives.
An article by Cook et al,2 on the evaluation of gastric cancers with computed tomography (CT), contains two conclusions unsupported by the data. The authors recommend "a continued role for celiotomy in managing gastric adenocarcinoma of the stomach since a significant percentage of patients whose disease is believed to be unresectable after CT have potentially curable lesions." A quick glance at their Table 2 shows that of 17 patients classified by CT as having stage III or IV disease (presumably predicting incurability), only three patients with stage IV disease had a "better" situation intraoperatively. We are not told the fate of these three patients, but to assume they all underwent resection and