Health Hazards In Mining The Files And Facilities

INTRODUCTION Radiation is one of the many agents occurring in our environment that is capable of causing cancers. We are all unavoidably exposed to natural background radiation. If we wish to exploit the beneficial uses of radiation in medicine, industry, and for the production of electricity, some further exposures to radiation are almost inevitable. One needs to ensure therefore, that the standards of radiation protection are safely derived and one would want to understand the mechanism by which radiation might cause cancer and genetic defects. In order to study the effects of radiation one can: (1) analyze human data, (2) perform animal experiments, and (3) do molecular level experiments. Indeed all three kinds of information are needed. This paper will concentrate on the steps required to obtain some of the necessary human data regarding health effects, where the concern is about the impact and consequences of long-term low-level exposures to an agent. Such investigations require some knowledge of work histories, dose histories, health "outcomes", and the personal identification of the individual involved. Three inter-related computer systems have been developed at Statistics Canada which have been designed to permit optimal use of a number of different records for our entire country for such health related research. The development of the Canadian Mortality Data Base, the initiation of the National Cancer Incidence Reporting System, and the development of new computer linkage techniques have helped reduce the cost and increase the scale and efficiency of automated follow-up to produce statistics of sickness or death associated with radiation and other carcinogenic agents in mines. These computer systems have already been implemented, and references will be made to studies currently being conducted using these files and facilities (e.g. a study of all Ontario miners, plus various Canadian uranium, fluorospar, salt and nickel miners). We will also look at the kinds of data that need to be collected now, to improve such studies in the future. DELAYED RISKS - THE STUDY SIZE AND COST Delayed effects on human health, as for example industrially caused cancer, can in general only be detected and measured by following-up the individuals to see what eventually becomes of them. What is not generally recognized is that the relatively low levels of individual risk, about which the public is often concerned, usually requires for their detection that very large numbers of "exposed" and "control" individuals (e.g. 10,000 to 100,000 or more) be followed over a period of two or three decades to determine when they die, what they die of, and whether they contracted cancer or some other disease of special concern. Thus, it is frequently exceedingly difficult to make such investigations cost-effective so that they will be undertaken at all, and as a result very real risks to health can remain undetected or unquantified. THE MANUAL PROCEDURES Follow-up of individuals by epidemiologists has until recently been a largely manual and clerical operation. It has used a diversity of source record files; local, regional, and national. Often the tracing of people has involved letters sent through the mails and visits to institutions, physicians, municipal offices, and former neighbours. Only thus could one find out whether the individuals were dead or alive. Such studies were necessarily small, or else very expensive. Death registrations have, by tradition, provided a valuable tool for the identification of harmful influences in the environment. However, with the increasing mobility in the population, death may often occur far away from the place of exposure to such an influence. Thus, no longer will simple manual searches in a single registry office suffice to inform the investigator concerning the deaths that may have occurred in a study population, especially when that population is large. In the past, manual follow-up to locate the relevant death registration normally required some prior knowledge of the province and year of death in question, so that the alphabetic indexes could he used to direct researchers to the appropriate bound volumes of registration forms. To search manually in this fashion for any large number of death registrations, without knowing in which year or province the deaths had taken place, or even whether they had yet occurred, would be impractical, since each year in Canada there are about 170,000 deaths. DEATH AND CANCER AS SPECIAL ENDPOINTS Much of recent effort has been focused on the organization of the "endpoint" files required to do long-term follow-up studies on a national scale, because that is a function which other institutions are unable to perform due to the confidentiality laws governing the use of such information. Outside organizations generally come to us with detailed "starting" point records which relate to some specific group requiring study. We carry out these epidemiological searches on a cost-recovery basis. The analytical interpretation of results is normally