Precision of Internal Dose Assessments

DOE-STD-1121-98

measurement results upon which these assessments can be based are usually slow in coming and highly

variable. Where the measurements can be obtained rapidly, it is often at a cost of analytical sensitivity,

which can raise the minimum dose detectable by bioassay.

The early clearance patterns in the first few days after intake are the most uncertain parts of the

biokinetic models, being highly affected by particle size, mode of intake, material transportability, and

individual person-specific metabolism (Traub and Robinson 1986). If an intake is quite minor, then these

issues are not particularly significant. This is because a conservative interpretation of early data using the

standard biokinetic models resulting in a small HE,50 (e.g., below 100 mrem) is not likely to cause any

major impact on classification of event.

High-energy photon-emitting radionuclides (e.g., fission and activation products such as 137Cs and

Co) are easily and quickly measured using whole body counter systems. Because incidents involving

these nuclides are usually small relative to the ALI, reasonably good early assessments of intake and dose

can be obtained with a high degree of confidence.

Such is not the case when dealing with plutonium and americium mixtures. These nuclides are

among the most difficult for which to provide confident early assessments. Errors in knowledge of the

mixture can lead to significant variations (factors of 2 to 10) in assessed doses. In vivo measurements are

relatively insensitive for plutonium mixtures. Likewise, early urine samples analyzed by a relatively

insensitive radiochemical procedure are not well-suited for dose assessments but may be very valuable for

initial determination of need for or efficacy of any dose reduction therapy. Large-volume urine samples

and fecal samples will provide better assessments of intake but will likely require several days to produce

results. The Hanford Site has developed an internal contamination incident response plan, contained in

the Hanford Internal Dosimetry Project Manual (Carbaugh et al. 1994), which specifically identifies the

capability of response as a function of time following intake and measurements made. For example, the

plan identifies the capability for various combinations of measurements following an aged weapons-grade

plutonium mixture inhalation to be as shown in Table XI. This table was derived for standard Hanford

dosimetry assumptions. Similar tables have been developed for other radionuclides and scenarios.

Preliminary assessments must be considered just that: It is not appropriate to place heavy reliance on

the actual magnitude of the dose in the first few days following a suspected intake. It would not be

unusual for a preliminary assessment of 10 or 20 rem CEDE derived from initial bioassay data for a

plutonium intake to ultimately be lowered to 1 rem CEDE based on long-term follow-up data.

9.3 PRECISION OF INTERNAL DOSE ASSESSMENTS

Interpreting bioassay data generally involves making many assumptions which can vary between

dosimetrists. Intercomparisons have been performed between DOE sites (Hui et al. 1994) and

internationally (Gibson et al. 1992). These comparisons have shown that ranges between 30% and 50%

of the mean value are not uncommon. In practical terms, this means that a factor of 2 to 3 variation

between dosimetrists is not unreasonable. Similar results were demonstrated by intercomparison of one

particular case (La Bone et al. 1992; La Bone and Kim 1993). A reassessment based on long-term data

increased the dose by a factor of 4 and also showed a factor of 2 variability around the mean assessment

of dosimetrists.

Knowledge about the relative precision (or imprecision) of internal dose assessments does not

relieve the site from making a precise conclusion about the dose to be assigned. It should be the

responsibility of the internal dosimetrist to decide on the best assessment of internal dose to be assigned

100