Experience from retrospective radon exposure estimations for individuals in a radon-epidemiological study using solid state nuclear track detectors.

The relation between an increased risk for lung cancer and exposure to indoor radon is assessed inepidemiological studies. Both the quality and reliability of smoking data and the radon exposuredata are of primary importance. Contemporary measurement of radon concentration in thedwellings of the individuals in a case-control study is traditionally used to assess the past radonhistory. These assessments contain large uncertainties since the contemporary measured radonconcentration might not be representative for the situation a long time ago. The measurement ofthe long-lived decay products from 222Rn remaining on indoor hard surfaces, such as glass,presents another possible way to assess the exposure to indoor radon.At the Swedish Radiation Protection Institute, a combination of two different solid state nucleartrack detectors has been developed to assess the 210Pb activity implanted in glass surfaces bymeasurement of 210Po alpha activity. This detector (a RETRO detector) is used in the Swedishradon epidemiological case-control study of non-smokers where the objectives with thesemeasurements are to provide an alternative estimate of individual radon exposure, to estimate theradon exposure when no other data are at hand and to evaluate the usefulness of RETROmeasurements.For this study we followed a pre-determined plan which gives guidance on how to choose theobjects, what information should be noted etc. The objects chosen for RETRO measurement werepersonal objects, which had been in the person's possession for more than 20 years. If possible,two different objects were measured for each individual. In addition, a contemporary radonmeasurement was undertaken during the three-month period during which the RETROmeasurement was made.Of the 419 persons contacted, we made at least one RETRO measurement for 344 of them, 9% hadno suitable object and 9% refused to allow measurements. A total of 576 different objects werefound and 568 were measured. For 225 persons we measured two personal objects that had been inthe same person's possession for more than 20 years. The standard deviation of the estimated meanradon concentration obtained from these two objects had a median value of 13 Bq/m indicating aprecision of exposure estimation of about 20 %.The correlation between 210Po surface activity measured earlier and mean values of measuredradon concentrations in a number of Swedish dwellings is applied to estimate the historicalaverage radon concentration. This average correlation factor seems also to be valid for themeasurements in the non-smoker epidemiological study.Measurement of surface implanted 210Po activity in personal objects can provide information onindividual radon exposure which is of value in a radon epidemiological study. Additionalinformation using room-models will improve the accuracy of the assessed exposure.