Purpose: To evaluate the evidence for the correlation between radiation exposure of the population following the Chernobyl accident and cancer mortality using statistical links.
Materials and methods: Data on deaths and causes of death for the period between 1993 and 2017 used in the study were taken from the National Radiation Epidemiological Registry (NRER). The number of deaths in males is 30771 persons including 5407 cancer deaths, the number of deaths in females – 29033 persons including 3472 cancer deaths. For analysis of statistical links between radiation exposure and causes of death data mining algorithms free of a priori statements on probabilistic distributions of doses and diagnoses were used. Tables of contingency of deaths in two dose groups: the group 0 – <0.014 Sv and the group 1 – ≥0.014 Sv; and in three age groups: group 0 – <17 years old, group 1 – 18–60 years old, group 2 – 60+ years old were used for analysis. About 90 % of individual effective doses were accumulated from 1986 till 1993. Mean effective dose is 0.024 Sv.
Results: For the population of four areas of the Russian Federation which are most contaminate after to accident on the CNPP significant association of all causes of death from cancer and for three-digit headings ICD-10 with an exposure dose was not found.
Statistically significant association between dose and pancreatic cancer (C25.9 ICD-10) and stomach cancer (С16.9) in males, stomach cancer (С16.9) and malignant neoplasms of bronchus and lungs (C34.9) in females was found. The association between dose and cancer death was demonstrated in 150–230 cases.
These diagnoses of causes of death first of all have to be a subject of more sensing and specific radiation and epidemiological analysis considering possible the confounding factors
Conclusion: The presented method is exploratory in nature and can be used to find a direction of research necessary for making more accurate evaluation of epidemiological evidence for the correlation between dose and effect. The method may be useful for evaluation of the association between radiation exposure and pathological response (death).
