Saccharin and Bladder Cancer-Reply

In the concluding paragraphs of our article, we sought to explain the divergent findings of our study and those of Howe et al (Lancet 2:578-581, 1977) on the basis of methodological distinctions. We pointed out that the Canadian investigators "apparently queried their patients on the use of tablet or drop sweeteners but not on powders." Dr Miller and his colleagues appear to concede this point in stating that "the use of powders is not specifically mentioned in the probing sentence on the questionnaire." They do not challenge our statement that "their interview omitted the ascertainment of calendar period in which sweeteners were used; thus, the analysis may have included subjects who used saccharin after the onset of cancer." Nor do they question our judgment that "their 'ever-or-never' analysis of saccharin use is subject to bias if patients and controls differed in the proportions reporting single, one-time, or occasional