Do childhood cancers result from prenatal x-rays?

Abstract An independent review of the published data from the Oxford Survey of Childhood Malignancies is presented. Publications of the Survey point out that the mothers of the children who died of cancer had a greater incidence of abdominal X-rays during pregnancy than the control group, and they attribute the difference in incidence to a cancer-causing effect of the X-rays exerted on the fetuses. Although this association of prenatal X-rays with cancer decreased for children born after 1957, the residual implied radiation effect is much larger than could be expected from the direct experience at Hiroshima and Nagasaki. The “crude dose-response” curves that have been published are shown to be faulty; they do not substantiate the existence of a casual relationship between prenatal X-rays and childhood cancer. Examination of the method used to calculate “relative risk” from X-rays shows that it requires an assumption that the sub-population of cancer cases has the same probability of being X-rayed as the sub-population of controls. The published data show that this is not the case and that, therefore, there is no basis for estimating any casual relationship. In fact, the sub-population of cancer cases is different from the control population in ways that indicate a greater need for medical care, with its associated X-ray procedures. Thus, the greater frequency of X-raying of the mothers of the children who died of cancer can be accounted for entirely by differences in sub-population characteristics.