Probability of causation for radiation-attributable cancer in the Indian population.

Once a cancer is diagnosed in an individual with a history of radiation exposure, it may be required to know the probability that radiation exposure was the cause of the disease. The National Institute of Health, U.S., has generated radio epidemiological tables giving probability of causation for various radiogenic cancers for the population of the U.S. In this paper, the probability of causation has been calculated for the Indian population using two models: (1) the original National Institute of Health model, and (2) direct use of Japanese constant relative risk coefficients for solid tumors. In both cases, new risk coefficients based on DS86 dosimetry and extended follow-up for 35 y have been used. Calculations with new coefficients, based on the National Institute of Health model have been extended to the American population and compared with the results for the Indian population. These values are generally higher for the Indian population than for Americans because of the lower baseline incidence rates in India. Probability of causation values based on the constant relative risk model are independent of population characteristics.