disease. The determinants of these risks are multiple. Exposure is a major determinant of infection risk, and it may be related to the work place, although it must be realized that health care workers may also face exposure in the communities in which they reside. Individual factors such as age, immune status, and genetic composition (possibly including race) are also important, especially for the risk of disease. These factors are not per se related to the workplace, but that does not mean they are not operative in the workplace.
There have been a number of well-done literature reviews of the occupational tuberculosis risk to health care workers published during the past decade (2–9). In general, these reviews emphasize the risk of tuberculous infection and do not deal with the subsequent risk of disease. Nor do they provide much information permitting one to compare the workplace risk to that incurred in the community in which health care workers reside. These reviews have documented that tuberculin conversion rates for American health care workers in the recent past were as high as 4–5 percent/year in many urban areas and perhaps twice that in some areas of New York City and for certain job situations with high exposures to aerosols of respiratory secretions. In nonurban areas, they were generally lower than 0.2 percent/year. With the implementation of enhanced infection control measures recommended by such advisory groups as CDC during the past decade, these rates have dropped to below 1 percent generally. However, outbreaks continue to occur from unrecognized sources, and with these outbreaks the tuberculin conversion rates among exposed employees may be as high as 50 percent.
This paper is based on a review of published literature. Searching was done using PubMed (MedLine) with a variety of topics relevant to tuberculosis and health care workers. Additional publications were selected from the bibliographies of published reports. Identified papers were then retrieved using the resources of the Health Sciences Library and Allen Memorial Library of Case Western Reserve University. Finally, a number of relevant papers were already present in the author’s personal library and reprint collection.
Data presented in this paper have been converted, when possible, from the form in which they were originally presented to percent per year. In a few reports, the original data are given as per 100 person-years, and this