Issues
Association rules for discovery relationship between mortality among Chernobyl liquidators and radiation dose
«Radiation and Risk», 2018, vol. 27, No. 1, pp.22-32
DOI: 10.21870/0131-3878-2018-27-1-22-32
Authors
Gorski A.I. – Lead. Researcher, C. Sc., Tech. A. Tsyb MRRC, Obninsk. Contacts: 4 Korolyov str., Obninsk, Kaluga region, Russia, 249036. Tel.: (484) 399-32-60; e-mail:
This email address is being protected from spambots. You need JavaScript enabled to view it.
.
Maksioutov M.A. – Head of Dep., C. Sc., Tech. A. Tsyb MRRC, Obninsk.
Tumanov K.А. – Head of Lab., C. Sc., Biol. A. Tsyb MRRC, Obninsk.
Kochergina E.V. – Head of Lab., C. Sc., Med. A. Tsyb MRRC, Obninsk.
Korelo A.M. – Senior Researcher. A. Tsyb MRRC, Obninsk.
Abstract
The aim of the study is to classify causes of death of the Chernobyl emergency accident workers (liquidators) in compliance with ICD-10 using statistical links. For these purposes we used mortality data (27840 cases) on Russian liquidators entered the Chernobyl zone in 1986-1987. The data were collected for the period from 1986 over 2014, average external dose of γ-radiation to a whole body is 0.110 Gy. To analyze statistical links between radiation doses 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 death cases in two dose groups, group 0, dose <0.1 Gy, and group 1, dose ≥0.1Gy, and two age groups, group 0, age <52 years, group 1, age ≥52 years, were used for analysis. Statistically significant positive association of radiation dose and death causes in similar age groups for disease in the Chapter C "Malignant Neoplasms" (age group 1) and the Chapter I "Diseases of the circulatory system" (age group 1). Statistically significant associations are in the blocs of categories C16 "Malignant neoplasm of stomach", C34 "Malignant neoplasm of bronchus and lung", I25 "Chronic ischaemic heart disease", all in the age group 1. The association is statistically significant for the diagnosis I25.9 "Chronic ischaemic heart disease, unspecified" in the age group 1. The positive odds ratio was found in the age group >52 years. This finding allows us to assume that the model of radiation risk is multiplicative because of the existence of positive correlation between radiation associated cancers and age, rather than additive. To determine correlation between causes of death and radiation dose more precisely, radiation risks should be assessed with account of results of radiation epidemiological studies of stochastic effects.
Key words
Chernobyl accident, ionizing radiation, doses, Chernobyl liquidators, mortality, data mining, ICD-10, tables of contingency, association rules, odds ratio, dose response.
References
1. United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). Sources and effects of ionizing radiation. UNSCEAR 2006 Report Vol. I, Annex A: Epidemiological studies of radiation and cancer. New York, United Nation, 2008.
2. Ivanov V., Tsyb A., Ivanov S., Pokrovsky V. Medical radiological consequences of the Chernobyl catastrophe in Russia: estimation of radiation risks. St. Petersburg, Nauka, 2004. 388 p.
3. Piatetsky-Shapiro G. Discovery, analysis and presentation of strong rules. Knowledge discovery in databases. Eds.: G. Piatetsky-Shapiro, W.J. Frawley. Cambridge, MA, AAAI/MIT Press, 1991. PP. 229-248.
4. Agrawal R., Imieliński T., Swami A. Mining association rules between sets of items in large databases. Proceedings of the 1993 ACM SIGMOD International conference on Management of data (SIGMOD'93). New York, 1993. PP. 207-216. doi 10.1145/170035.170072.
5. Hahsler M. A Probabilistic Comparison of Commonly Used Interest Measures for Association Rules, 2015. Available at: http://michael.hahsler.net/research/association_rules/measures.html (Accessed 16.10.2017).
6. Doddi S., Marathe A., Ravi S.S., Torney D.C. Discovery of association rules in medical data. Med. Inform. Internet Med., 2001, vol. 26, no. 1, pp. 25-33.
7. Ordonez C., Ezquerra N., Santana C.A. Constraining and summarizing association rules in medical data. Knowledge Information System, 2006, vol. 9, no. 3, pp. 259-283.
8. Jensen S. Mining medical data for predictive and sequential patterns. Proceedings of the Fifth European Conference on Principles and Practice of Knowledge Discovery in Databases. Freiburg, Germany, September 3-5, 2001. P. 1-10.
9. Shantakumar B.P., Kumaraswamy Y.S. Extraction of significant patterns from heart disease warehouses for heart attack prediction. International Journal of Computer Science and Network Security (IJCSNS), 2009, vol. 9, no. 2, pp. 228-235.
10. International Statistical Classification of Diseases and Related Health, 10th revision (ICD-10). Vol. 1 (Part 1). Geneva, WHO, 1995. 696 p. (In Russian).
11. Software «Statistics». Available at: http://www.statsoft.ru (Accessed 16.10.2017).
12. Mietenen O.S. Confounding and effect modification. Am. J. Epidemiol., 1974, vol. 100, pp. 350-353.
13. Gorsky А.I., Maksioutov М.А., Tumanov K.A., Shchukina N.V., Chekin S.Yu., Ivanov V.K. Non-parametric analysis of radiation risks of mortality among Chernobyl clean-up workers. Radiacionnaja biologija. Radiojekologija – Radiation Biology. Radioecology, 2016, vol. 56, no. 2, pp. 138-148. (In Russian).
14. Ivanov V.K., Tsyb A.F., Gorsky A.O., Maksioutov M.A., Chekin S.Yu, Petrov A.V., Tumanov K.A., Kashcheev V.V. Cancer morbidity and mortality among Chernobyl emergency workers: estimation of radiation risks. Radiacionnaja biologija. Radiojekologija – Radiation Biology. Radioecology, 2006, vol. 46, no. 2, pp. 159-166. (In Russian).
15. Ivanov V.K., Maksioutov M.A., Chekin S.Yu., Petrov A.V., Biryukov A.P., Kruglova Z.G., Matyash V.A., Tsyb A.F., Manton K.G., Kravchenko J.S. The risk of radiation-induced cerebrovascular disease in Chernobyl emergency workers. Health Phys., 2006, vol. 90, no. 3, pp. 199-207.