EPR dose reconstruction for Russian nuclear workers.

The present paper describes results of the first independent evaluation of the accuracy of the dose-reconstruction technique by electron paramagnetic resonance using tooth enamel. Each of twenty-four teeth donated by Mayak nuclear workers with known occupational radiation exposure histories was cut into two parts so that each tooth could be shared for blind electron paramagnetic resonance examination by at least two of the four laboratories participating in the study. The mean difference (+/- SD) between electron paramagnetic resonance estimates of the paired samples of each tooth shared by the two laboratories in best agreement was 0.02 +/- 0.15 Gy. Mayak workers can be classified into two groups: modern workers (after 1961) having reliable official dose information and earlier nuclear workers whose dose information is less reliable because they were exposed mainly before 1961. Film badges did not contain filters until 1954. Doses in this earlier group are much higher (up to 5 Gy). Comparison of the electron paramagnetic resonance results with tooth doses calculated from official film-badge doses showed a close agreement for the first group, whereas in the second group, official doses appeared to be slightly higher than the electron paramagnetic resonance doses. The results suggested a possibility that the official doses were somewhat overestimated among the high-dose-exposed workers. Consequently, cancer risks derived from this high-dose group might tend to be slightly underestimated.