Reply to Jargin’s comments: solid cancer increase among Chernobyl liquidators—alternative explanation

Abstract

We cannot agree with commentaries of Dr. Jargin on our article (Kashcheev et al. 2015). In these commentaries, Dr. Jargin theorizes about the quality of health examination in Russia and provides his own ‘‘alternative’’ the explanation of the magnitude of SIR (SMR) of solid cancer among the cohort of emergency workers (liquidators). On this basis, he makes wrong conclusions on radiation risks among liquidators. Note that radiation risks (ERR/Gy and RR) of solid cancer incidence and mortality estimated in our study do not depend on SIR and/or SMR. Dr. Jargin makes assumption that ‘‘it is hard to believe that surveillance bias was not operating in the study.’’ In the study (Kashcheev et al. 2015), we used two types of cohort epidemiological methods: external and internal comparisons (Breslow and Day 1987). Comparison with an external standard—solid cancer incidence and mortality among emergency workers were compared with spontaneous incidence and mortality among males of Russia based on the analysis of the standardized incidence ratio (SIR) and the standardized mortality ratio (SMR) (Eq. 1 and Eq. 2 in Statistical method section, respectively). Comparison with an internal control group—the excess relative risk (ERR/Gy) and the relative risk (RR) of solid cancer incidence and mortality were estimated (Eq. 3 and Eq. 5 in Statistical method section, respectively). In external comparison, the SIR (SMR) values were used for estimating the magnitude of this ‘‘surveillance bias.’’ Dr. Jargin notes correlation between variability of SIR values for cancer among liquidators and the healthcare quality in different geographical areas. We analyze the geographical variability in the first paragraph of the Discussion section (Kashcheev et al. 2015), and it is taken into account in estimating relative and excess relative risks (Eq. 3 and Eq. 5 in Statistical method section, respectively). There is no correlation between difference in SIR estimates and life expectancy, which Dr. Jargin suggests. SIR and SMR are calculated with the use of weight of personyears for 15 groups of attained age (Statistical method section). When internal comparison is made (radiation risk estimates), surveillance bias ‘‘was not operating,’’ because all emergency workers (liquidators) of selected cohorts are covered by the similar level of health examination, which does not depend on geographical areas (Cohort description section of the article). According to personal information accumulated in the Russian National Medical and Dosimetric Registry (RNMDR), frequency and quality of compulsory health examination do not depend on radiation dose. Therefore, we cannot accept Dr. Jardin’s speculation about dose-dependent difference in the quality of medical surveillance. Evaluating dose–response relationship is one of the Bradford Hill criteria to be applied in testing whether or not an association is judged to be casual (Hill 1965). The aim of our study was to determine whether this dose–response relationship exists in the cohort of Russian liquidators or not. We followed principles of the current radiological protection system recommended by international scientific committees and organizations (ICRP 2007; UNSCEAR 2011; WHO 2013). Both high and low doses may cause stochastic effects (cancer or heritable effects), which may be observed as a statistically detectable increase in the & V. Ivanov nrer@obninsk.com