Abstract
In the past, radiation exposures at or below the minimum detectable level (MDL) were generally considered insignificant to any health effects. More recently, these relatively low and imprecisely measured exposures are being scrutinized more closely to determine if exposure at these levels may be responsible for measurable excess illness or deaths. Using photographic film dosimeters to determine external gamma radiation at the U.S. Atomic Energy Commission's operations during the years 1942 to 1955, the MDL was typically considered to be 300 to 500 microsieverts (μSv). At some facilities, many Workers' dosimeters were read weekly and, during some periods, dosimeter readings from film densities less than the MDL were set to zero. Depending on the frequency of monitoring and the magnitude of the actual film density, considerable annual exposure may have been unrecorded. Health physics monitoring data and calibration data were examined to determine appropriate modeling parameters for estimating the fraction of actual exposure that would be unrecorded when an MDL rule (setting readings less than the MDL to zero) was applied. Frequency distributions of readings were determined for workers at various reported annual dose levels, categorized by estimated mean weekly exposure. Assuming a lognormal distribution of actual readings and an MDL of 300 μSv, from 4000 to 7500 μSv additional annual exposure was probably unrecorded for workers with recorded annual exposures of 500 to 12,000 μSv. It is concluded that: (1) levels of radiation exposure that now may be considered significant may have been unrecorded; (2) the total unrecorded exposure is dependent on the frequency of monitoring, the MDL applied, and the variability of readings; and (3) the maximum unrecorded exposures would have occurred in employees with actual mean weekly exposures between 30 and 120 percent of the MDL.