In reviewing their data, McCray and colleagues concluded that the risk of new tuberculous infection for most health care workers was not substantially greater in the workplace than in the community in which they resided. The annual risks calculated from the data of McCray and colleagues are higher than those estimated and previously presented in this report for general populations. Their data were based on prospectively collected and currently reported information available in more specific categories with better demographic stratification than the data used for the author’s estimations of ARIs. However, there may be a selection bias introduced by the choice of sites for the study of McCray and colleagues.
The CDC data suggest that the tuberculosis risks of health care workers closely parallel those for the communities in which they reside. This does not mean that transmission of tuberculosis infection does not occur in the health care-related workplace. It simply means that the occupational risk is not great compared with the community risk.
There are no studies available allowing one to estimate the risks of tuberculous disease and mortality in M. tuberculosis-infected health care workers per se. One must generalize from what is known about these risks in the general population. This risk has frequently been stated to be about 10 percent over the life of the infected individual, but the available data suggest that it is closer to 5 percent, with about half of the risk occurring in the first 1 to 3 years after infection. In fact, the risk of tuberculosis for infected health care workers should be less than that for other persons because they work in circumstances that are optimal for monitoring of tuberculin test conversion, for implementation of therapy of latent infection, and for education and orientation concerning the importance of this form of therapy. Isoniazid treatment of latent tuberculous infection has been shown to reduce the risk of disease by about 60 percent (82). Similarly, the mortality risk should be low because health care workers should have prompt access to detection and therapy of disease. In considering these risks, it is important to distinguish between those in immunocompetent persons and those in immunocompromised individuals.
J.Arthur Myers and coworkers traced University of Minnesota medical students who were tuberculin-positive at medical school entry or who