Uncertainty of Estimated Radiation Doses Among Atomic Bomb Survivors: Using Multiple Sources of Information on Errors in the Main Exposure Variable of Health Effects Studies

• Activity Number: 233
• Type: Topic Contributed
• Date/Time: Tuesday, July 31, 2007 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Harry Cullings*+ and Eric J. Grant and Sachiyo Funamoto and Sachiko Teranishi
• Companies: Radiation Effects Research Foundation and Radiation Effects Research Foundation and Radiation Effects Research Foundation and Radiation Effects Research Foundation
• Keywords: exposure uncertainty ; radiation dosimetry ; biodosimetry ; dose error

Abstract

Studies of atomic bomb survivors in Hiroshima and Nagasaki use doses estimated by a dosimetry system as the main exposure variable; these inherit uncertainty from many sources. There is no subset of survivors with dose estimates suitable for validation, but there are many sources of partial information about dose uncertainty. Progress has been made in the past with methods for dealing with dose error based on assumed joint distributions of true and estimated doses. This talk will focus on sources of ancillary information that could improve uncertainty estimates, and their relation to error models. Examples include reliability of recall in re-interviewed survivors, population density, spatial proximity in families, influence of distance and shielding uncertainty based on physical propagation of error, biodosimetry, environmental measurements, and data on acute signs of radiation exposure.