Royal Statistical Society Meeting on Cancer Near Nuclear Installations

Abstract

During the last few years there have been reports of excesses of childhood cancer -in particular leukaemia-around nuclear installations in Britain, most notably the reprocessing plants at Sellafield in Cumbria and Dounreay in northern Scotland. The issue of whether excesses of cancer are occurring in small geographical areas around nuclear installations and, if so, their likely cause is not only one of great public interest but is also very much a statistical one. However, until recently statisticians, especially those working on methodological aspects of the analysis of spatial data, have not taken a leading role in attempting to resolve the issue. Consequently the General Applications Section of the Society, in conjunction with the Medical Section, organized a half-day Ordinary Meeting with the title Cancer near Nuclear Installations on May 17th, 1989, the proceedings of which are published here. The purpose of the meeting was to bring together those with expertise in a variety of relevant fields. It began with a series of four papers describing the background to the subject. Professor Martin Gardner, who was a member of Sir Douglas Black's advisory group that investigated reports of excess childhood cancer around Sellafield and who is now a member of the Committee on Medical Aspects of Radiation in the Environment (COMARE), reviewed reports of increased childhood cancer rates near UK nuclear installations; Dr Tom Wheldon, who is also a member of COMARE, described current methods of assessing radiation-related risks; Sir Richard Doll summarized the epidemiology of childhood leukaemia; and Dr Michael Hills and Dr Freda Alexander reviewed relevant statistical methods. These papers were followed by an invited discussion made up of five contributions from statisticians with substantial methodological knowledge of the analysis of spatial data. Finally there was a contributed discussion in the usual format for Ordinary Meetings of the Society. New data sets making extensive use of small area statistics are currently being compiled by, for example, the Childhood Cancer Research Group in Oxford and the Small Area Health Statistics Unit in London. Also, case-control studies are being conducted near several UK nuclear installations, and studies are now being undertaken around installations overseas. It is hoped that the papers and discussion from this meeting will prove helpful to statisticians and others who become involved in the analysis of data on this topic.