Occupational experience

Abstract

The use of occupational information to find out about causation of disease is a practice of long standing. The working population is partly selected and partly self-selected. It cannot fully represent the general population since it excludes children, some disabled people or those susceptible to disease, and some women. Exposure to a contaminant at work in most cases is limited to 8 h a day. However, in many cases the exposures are much higher than those in the outside environment and, if total dose is the main damaging factor, the hazard is consequently higher. For this reason the first place to seek information on environmental impact will often be the workplace. Even if the hazard is not large, correlation may be possible if the associated lesion is unusual. Most difficulties occur when a relatively common condition is involved, and when exposures to the possible agent are so low that the population required to show significant statis­tical variation will be unreasonably high. In this situation, individual attribution to cause may be unrealistic against the background of naturally occurring conditions. Earlier detection by biological indicators may be suggestive but too non-specific to be decisive. Additionally, secondary contamination (neighbourhood cases) may by its geographical distribution enable hazard identification to be made and dose-response estimated. The risks associated with new substances have led to the introduction of proposals for a notification scheme. How this is to be related to specific follow-up and analysis is the big epidemiological problem of the time. Mortality studies are too slow, though large-scale ones provide an ultimate test of any theory of dose and response.