Kellerer, A. M. and Walsh, L. Solid Cancer Risk Coefficient for Fast Neutrons in Terms of Effective Dose. Radiat. Res. 158, 61–68 (2002).

Cancer mortality risk coefficients for neutrons have recently been assessed by a procedure that postulates for the neutrons a linear dose dependence, invokes the excess risk of the A-bomb survivors at a γ-ray dose D₁ of 1 Gy, and assumes a neutron RBE as a function of D₁ between 20 and 50. The excess relative risk (ERR) of 0.008/mGy has been obtained for R₁ = 20 and 0.016/mGy for R₁ = 50. To compare these results to the current ICRP nominal risk coefficient for solid cancer mortality (0.045/Sv for a population of all ages; 0.036/Sv for a working population), the ERR is translated into lifetime attributable risk and is then related to effective dose. The conversion is not trivial, because the neutron effective dose has been defined by ICRP not as a weighted genuine neutron dose (neutron kerma), but as a weighted dose that includes the dose from γ rays that are induced by neutrons in the body. If this is accounted for, the solid cancer mortality risk for a working population is found to agree with the ICRP nominal risk coefficient for neutrons in their most effective energy range, 0.2 MeV to 0.5 MeV. In radiation protection practice, there is an added level of safety, because the effective dose, E, is—for monitoring purposes—assessed in terms of the operational quantity H*, which overestimates E substantially for neutrons between 0.01 MeV and 2 MeV.