Radon retrospective dosimetry for the case-control study in lazio region: a feasibility study

^ Department of Experimental Physics, University College Dublin, Dublin 4, IrelandCase-control studies on lung cancer and radon in dwellings have been conducted in North Americaand Europe, including one study in Lazio region, in Central Italy. Great attention is generally devotedto the reduction in the uncertainty in radon exposure assessment because it can give rise to relevantuncertainty in the estimated risk. One of the main potential sources of uncertainty is the use ofcontemporary radon concentration measurements as estimators of past concentration values. This isbecause significant variations of radon concentration could be produced by changes in the buildingand dwelling structures or in the living habits. Another method to estimate the radon exposure of thesubjects is the recently developed retrospective dosimetry on glass objects. In particular, this techniquecould be effectively used as a quality control for exposure assessments obtained from contemporaryradon concentration measurements. In order to evaluate the feasibility of such a technique for theLazio case-control study, a study has been carried out in a sample of 26 dwellings in Rome, havingboth low and high radon concentration values (from 28 Bq/m3 to 623 Bq/m3). Retrospective detectorshave been exposed from two to three months on 50 glass objects in bedrooms and living rooms inthese dwellings. The retrospective detectors are actually a combination of two alpha track detectors(CR-39 and LR-115) with different alpha energy registration characteristics, which are used side byside in order to separate the signal from surface 210Po activity and that from the energy degraded alphaactivity of the bulk glass. The average radon concentrations estimated from the measured 210Po surfaceactivities on glass objects are compared with the current radon concentration measured for one yearwith LR-115 alpha track detectors. For most of the glasses the ratio R (= contemporary Rn Conc. /Retrospective Rn Conc.) is within or very close to the range 1/3 3 considered acceptable, taking intoaccount the uncertainties of the techniques. The correlation factor between the two sets of data is quitelow (r=0.42), however it is significantly affected by the six most extreme values, three of which comesfrom a single room full of electronic equipment. If these data are removed r=0.67, which is similar toresults obtained in other validation studies. Correlation does not increase if retrospective estimationsare corrected for aerosol production associated with smoking.