Читать книгу Industrial and Medical Nuclear Accidents - Jean-Claude Amiard - Страница 68

Several authors [BRO 82, DEW 87] point out that populations near Three Mile Island (TMI) have developed mental trauma due to anxiety created by unnecessary and careless publicity.

The excess percentage of morbidity caused by the accident for the period 1981–1984 (after a latency of 2 years) was estimated per unit dose (associated with the standard deviation). These percentages were 0.020 ± 0.012 for all cancers, 0.082 ± 0.032 for lung cancer and 0.116 ± 0.067 for leukemias. Adjustments for socio-economic variables increase the percentage estimates to 0.034 ± 0.013, 0.103 ± 0.035 and 0.139 ± 0.073, respectively, for all cancers, lung cancers and leukemias [WIN 97].

According to Talbott et al. [TAL 00], standardized mortality ratios are significantly raised for men and women (SMR 109 and 118, respectively). They also found a significant linear trend between breast cancer risk in women and increasing levels of exposure associated with the TMI accident (p = 0.02). In a subsequent study, the same authors [TAL 03] were more reserved about cancer and radioactivity relationships caused by the TMI accident than in their first article.

In a synthesis conducted 25 years after the accident, Osif et al. [OSI 04] estimated that the health effects resulting from the radiation emitted by the reactor accident were minimal. The only significant effects were on the mental health of local populations. However, the studies carried out only concern a part of the population and over a short period of time. Some effects may appear years or even decades later. For example, 30 years after the TMI accident, an increased incidence of thyroid cancer was observed in counties south of the plant and in high-risk age groups. Average incidence rates between 1990 and 2009 were higher than expected in the counties of York, Lancaster, Adams and Chester [LEV 13]. Despite these findings, a direct correlation with the accident remains uncertain, as incidence rates may coincide with other factors, and the original data were limited [LEV 13].