An approach for conducting epidemiological surveillance in industry

There are three unresolved issues in epidemiological surveillance: 1) should surveillance be hypothesis generating, 2) how much effort should be put into the design and data collection of a surveillance project, and 3) can useful surveillance be done with questionable data? It is concluded that epidemiological surveillance should be approached as an epidemiology study that examines the potential association between specific exposure and disease. This would include 1) a feasibility study to determine if surveillance on a particular subject is possible, 2) a detailed protocol, and 3) plans for data reduction, interpretation of findings, and worker communication. Epidemiological surveillance must be well planned and resourced to be used as a screening mechanism for occupational illness or injury excess, monitoring the health of workers exposed to specific substances, or setting priorities for future special studies.