Conversely, malignant or hereditary effects cannot be unequivocally attrib­uted to radiation exposure on individual bases for reasons of counterfactual conditionality. This is because radiation exposure is not the only possible cause of these types of effects and, at present, no biomarkers are available for these effects that are specific to radiation exposure.

However, while the occurrence of malignant effects (or of hereditable effects in the descendants of those exposed) cannot be unequivocally attrib­utable to radiation on an individual basis, an increased incidence of these effects in a population can theoretically be attributed to radiation on a col­lective basis. Collective attribution can be established through epidemio­logical analysis, under the following conditions:

• The number of cases of the effect in the exposed population should be sufficient to overcome the inherent aleatory uncertainties of epidemio­logical analyses.

• In addition, the increase in the collective prevalence of the effect in the exposed population is properly attested by a qualified radio-epidemio­logical procedure.

In situations of chronic exposures at levels similar to those arising from normal operations of NPPs, the expected number of additional cases of malignancies in a commensurate population for an epidemiological study would be so low that attribution is unattainable either individually or collectively.

Thus, while increased incidences of malignancies and hereditary effects might theoretically occur in populations exposed to NPPs, since it is not feasible to obtain unequivocal scientific evidence of their occurrence, they therefore should neither be deemed attributable nor be used prospectively in notional projections of radiation harm. Moreover, hereditary effects cannot at present be attributed to radiation exposure, even at high doses, because the fluctuation in the normal incidence of these effects is likely to be so much larger than any expected radiation-related increase in the incidence.

It is to be noted, however, that occasionally individual attribution can nevertheless be ostensible, namely, apparently factual, even if not necessar­ily so; this may be the case when:

• the ‘background’ incidence of the effect is low, and

• the radio-sensitivity of the effect is high.

A typical example of ostensible individual attributibility is the case of fol­licular thyroid cancer in children exposed to relatively high thyroid doses such as those incurred after the Chernobyl accident.