As discussed in Chapter 3, courts have directed OSHA to undertake quantitative risk assessments to justify its standards. Such risk assessments are often difficult because relevant data about the full extent of a workplace hazard and the consequences of control measures are very limited. These difficulties are present in full measure for the assessment of the 1997 proposed rule on occupational tuberculosis.
In the 1997 proposed rule, OSHA presented a quantitative risk assessment that estimated the number of infections, cases of disease, and deaths due to tuberculosis that would be averted by adoption of the rule. OSHA staff had four experts review an earlier version of the risk assessment, and they made some revisions on the basis of the reviewers’ comments. In preparation for the issuing of a final standard on occupational tuberculosis, OSHA staff have again revised and updated their estimates. This new analysis was not, however, available to the committee pending publication of the final standard. Therefore, the following comments necessarily apply to the earlier analysis included in the 1997 proposed rule.
Although OSHA published the proposed rule in 1997, much of the data on which it relied were several years older (e.g., a 1994 Washington State survey, 1991 data from Jackson Memorial Hospital in Miami, and a 1984–1985 North Carolina study). As summarized in this chapter, tuberculosis cases, case rates, and deaths have been declining since 1993, and recent studies also suggest low levels of occupational transmission of M. tuberculosis. The changing epidemiology of tuberculosis reflects both community and workplace tuberculosis control measures. Given this change, it is not surprising that the assessment presented with the 1997 proposed rule is outdated and that OSHA has revised it. (Again, the revised analysis was not available to the committee.)
In 1997, OSHA defined infection with M. tuberculosis as the “material impairment of health.” It did so on the basis of both the potential for latent infection to progress to active disease (which is discussed further below) and the risk for adverse health effects from treatment of the infection. As described earlier, although treatment is not risk free and individuals offered treatment should be informed of both benefits and risks, recent data suggest that the risk of liver damage from carefully monitored treatment of latent infection using isoniazid is quite low and is less than that described in the proposed rule.
Estimation of levels of tuberculosis infection and potential reductions in such infections as a result of an OSHA standard is particularly