central estimate (50th percentile) is 18.25 That confidence interval encompasses an estimate of uncertainty by the US Environmental Protection Agency (EPA, 1999). On the basis of the uncertainty estimated by Kocher et al., a credible upper bound of the biological effectiveness of alpha particles, as represented by the 95th percentile, is about 76, or a factor of nearly 4 greater than the value 20 used in dose reconstructions. An uncertainty of such magnitude clearly is important.
 Dose coefficients for inhalation of plutonium used in dose reconstructions may underestimate doses to organs of the GI tract by a substantial amount in scenarios in which an appreciable fraction of inhaled materials are respirable (AMAD, 1 μm).
In dose reconstructions for atomic veterans, doses to organs of the GI tract due to inhalation of radionuclides usually are calculated by using dose coefficients for large particles (AMAD, 20 μm). That choice is made because, on the basis of dose coefficients for inhalation and ingestion used in the NTPR program, an assumption of large particles generally results in higher estimates of dose to these organs than an assumption of respirable particles (AMAD, 1 μm); see Section IV.C.2.2.1 and dose coefficients for the wall of the lower large intestine based on ICRP Publication 30 (ICRP, 1979a) given in Tables V.C.1 through V.C.4.
On the basis of dose coefficients for inhalation and ingestion of radionuclides by adult workers currently recommended by ICRP (1994a; 2002), an assumption of inhalation of large particles still results in higher estimates of dose to organs of the GI tract in most cases (see Tables V.C.1 through V.C.4). However, insoluble plutonium is an exception. In this case, current ingestion dose coefficients for organs of the GI tract are little changed from the values used in dose reconstructions, but inhalation dose coefficients for these organs, assuming an AMAD of 1 μm, are a factor of about 10 higher in the current recommendations. As a result of that increase, the usual assumption of large particles would result in an underestimate of dose to organs of the GI tract from inhalation of insoluble plutonium by a factor of about 5 in scenarios in which inhalation of respirable particles is likely. The importance of this underestimation of dose depends on intakes of plutonium relative to intakes of other radionuclides and the magnitude of the dose.
The increase in dose coefficients for organs of the GI tract from inhalation of insoluble plutonium and the counterintuitive result that the dose to these organs from inhalation is higher than the dose from ingestion of insoluble plutonium is due to a significant change in the biokinetic model for systemic plutonium. In the model used in the NTPR program (ICRP, 1979a), all plutonium absorbed into