National Cancer Institute study: evaluation of computed tomography in the diagnosis of intracranial neoplasms. II. Was randomization necessary?

The final results of the National Cancer Institute study were evaluated to determine whether there were any differences in the comparative ability to diagnose intracranial tumors depending upon the order in which computed tomography (CT) and radionuclide scanning (RN), the two basic modalities, were performed. Only those patients who were examined by noncontrast CT and RN (with an interval of one or more days between the tests) were studied for this paper. There were 684 patients who underwent CT before RN, and 959 who underwent RN before CT. It was found that any effects of order, if they exist at all, were sufficiently unimportant that they did not change the results of the study. It is concluded that, for this study, randomization was not necessary.