Technical Report No. 7-90
An investigation of random errors in the DS86 dosimetry using data on chromosome aberrations and severe epilation
Sposto R, Stram DO, Awa AAEditor’s note: A publication based on this report was published in Radiat Res 128:157-69, 1991.
Summary
An analysis of the proportion of cells with chromosome aberrations (CA) in cultured blood lymphocytes from A-bomb survivors in Hiroshima and Nagasaki reveals 1) that the dose-response relationship using DS86 assigned dose is significantly steeper in the subsample of individuals who reported severe epilation after the bombings than among those who did not report severe epilation, and 2) there is substantially more variation among individuals exposed at higher doses in the proportion of cells with CA than would be expected if there were no heterogeneity in dose response among individuals. Since these effects are due to some extent to random errors in the DS86 dose assignments, one might envision using them to estimate the magnitude of random errors in the DS86 dosimetry. However, such an attempt is complicated by the fact that these effects could also be due to individual differences in sensitivity to radiation, so that at a given level of exposure, individuals who are more sensitive to induction of severe epilation are also more prone to induction of CA.
In this paper, working within a class of dosimetry error models which will be described below, we estimated the magnitude of random dosimetry errors which would be required to account for all of the difference in the observed dose response between persons who did and did not report severe epilation, and for all of the overdispersion in the proportion of cells with CA, under the assumption that random dosimetry error is the only cause of these effects. By performing analyses of these data over a range of dosimetry error models, we conclude that random dosimetry errors in the range 45% to 50% of true dose are necessary to explain completely the difference in dose response between the two epilation groups. This level of dosimetry error also accounts for the overdispersion in the proportion of cells with CA. Random dosimetry errors as small as 40% are consistent with the assumption that there are no differences in radiation sensitivity among individuals. We compare these dosimetry error figures with those obtained independently by others. We also discuss evidence that the contribution of differences in radiation sensitivity may be quite small, so that random dosimetry errors may be the most likely cause of these effects. Additional analyses of the type we have undertaken, using different endpoints, could lead to a better understanding of the amount of random dosimetry error associated with the DS86 dosimetry, and thus more accurate estimation of dose-response relationships based on RERF data.
An analysis of the proportion of cells with chromosome aberrations (CA) in cultured blood lymphocytes from A-bomb survivors in Hiroshima and Nagasaki reveals 1) that the dose-response relationship using DS86 assigned dose is significantly steeper in the subsample of individuals who reported severe epilation after the bombings than among those who did not report severe epilation, and 2) there is substantially more variation among individuals exposed at higher doses in the proportion of cells with CA than would be expected if there were no heterogeneity in dose response among individuals. Since these effects are due to some extent to random errors in the DS86 dose assignments, one might envision using them to estimate the magnitude of random errors in the DS86 dosimetry. However, such an attempt is complicated by the fact that these effects could also be due to individual differences in sensitivity to radiation, so that at a given level of exposure, individuals who are more sensitive to induction of severe epilation are also more prone to induction of CA.
In this paper, working within a class of dosimetry error models which will be described below, we estimated the magnitude of random dosimetry errors which would be required to account for all of the difference in the observed dose response between persons who did and did not report severe epilation, and for all of the overdispersion in the proportion of cells with CA, under the assumption that random dosimetry error is the only cause of these effects. By performing analyses of these data over a range of dosimetry error models, we conclude that random dosimetry errors in the range 45% to 50% of true dose are necessary to explain completely the difference in dose response between the two epilation groups. This level of dosimetry error also accounts for the overdispersion in the proportion of cells with CA. Random dosimetry errors as small as 40% are consistent with the assumption that there are no differences in radiation sensitivity among individuals. We compare these dosimetry error figures with those obtained independently by others. We also discuss evidence that the contribution of differences in radiation sensitivity may be quite small, so that random dosimetry errors may be the most likely cause of these effects. Additional analyses of the type we have undertaken, using different endpoints, could lead to a better understanding of the amount of random dosimetry error associated with the DS86 dosimetry, and thus more accurate estimation of dose-response relationships based on RERF data.