Abstract
The effect of screening on cancer is often problematic, as it may lead to overestimation of cancer incidence rates and overtreatment. Following the 2011 Fukushima Daiichi nuclear disaster, a thyroid ultrasound examination program (Thyroid Screening Program) was established for residents aged ≤18 years in Fukushima Prefecture, Japan. During the first three years of the screening, 300,476 children were tested, and 113 thyroid cancer cases were either identified or suspected. As no equivalent data on occult thyroid cancer among children are available elsewhere, this result brought much confusion on the interpretation of the data, even among scientists.
An article by Tsuda et al. (1) published in Epidemiology in October 2015 clearly illustrates this problem. The article describes the results of thyroid cancer screening in Fukushima after the 2011 nuclear disaster by making two types of comparisons. For an “external comparison,” a national annual incidence rate estimation based on hospital data was used as a control, and the authors concluded that the estimated incidence rate of thyroid cancer among the children of Fukushima is approximately 30 times higher than that of other areas. For an “internal comparison,” the authors used administrative districts and categorized each district into three areas: most, intermediate, and least contaminated area. There were no significant differences between these areas.
Many researchers have already posted comments on the potentially deleterious consequences of this screening effect. As illustrated by another study (2), the thyroid cancer cases identified among children in Fukushima is 20–30 times higher than would be expected without screening. However, based on estimated radiation exposure doses among the cancer cases and from dose–response gradients obtained from a cohort study in Chernobyl, the authors of this study suggested the increase is mainly due to overdiagnosis (2).
Additionally, according to the report by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) (3), the highest district average absorbed dose by the thyroid of a one-year-old infant in the first year was estimated to be about 50 mGy (effective dose of 7.5 mSv), and it was found in Iwaki City. However, in the “internal comparison” performed by Tsuda et al. (1), Iwaki City was assumed as a “least contaminated area.” More inconsistent was that even though the authors suggest that the “prevalence odds ratio (in the central middle district) compared with the reference district was 2.6,” this odds ratio showed no statistical significance [confidence interval 0.99–7.0]. As any correlation between the estimated radiation exposure levels and incident rate ratios of thyroid cancer cannot be assumed without statistical significance, this misleading description, intentionally or not, need to be corrected or deleted. Moreover, as about two-thirds of the exposure to 131I was due to ingestion (3), the exposure levels vary by individual food intake rather than by air-dose rates of the various residential areas. Therefore, it seems hardly valid to use administrative districts for estimation of contamination level by radioactive iodine.
In addition to these scientific considerations, we also claim that such a misleading description could become a burden of irrational stigmatization of children in Fukushima. For example, healthcare professionals in Fukushima, including us, provide lectures on radiation to school children, and often receive comments of concern in the questionnaires after the lectures, such as “we will all get cancer in the end,” or “I have resigned myself to being single,” among others. As increased psychological distress has been reported elsewhere across Fukushima (4), publishing tendentious data can aggravate negative emotions and contribute to a deterioration of the health status of children in the disaster area.
We do respect the authors’ considerable effort to estimate the incidence rate of thyroid cancer based on the currently available published data, in which personal information of each examinee is unavailable. However, interpretation of such data without knowing the background of each patient, such as whether and where they were evacuated after the disaster or how much their iodine intake was, might skew the vision of the real world. Considering that the findings may be potentially harmful to children in Fukushima, as they may fuel stigmatization, we strongly recommend that scientists should provide rational validity of their comparisons, as well as scientifically accurate interpretation of the results.