was not associated with patient contact or occupational category, so analysts concluded that community rather than occupational exposure was more likely.
In a national survey that focused on infection control measures, responding hospitals showed overall annual skin test conversion rates of 0.6 to 0.7 percent for 1989 through 1992 (Sinkowitz et al., 1996). Higher conversion rates were found among those involved in bronchoscopy (3.7 percent) and respiratory therapy (1.0 percent). No information was collected on employee demographic characteristics or on the details of the testing procedures used by the facilities.
Another study reporting higher rates of skin test conversions for respiratory therapists involved a military medical center in Maryland (Ball and Van Wey, 1997). Annual skin test conversion rates, which ranged between 0.4 and 2.6 percent across the occupational categories identified, did not differ significantly for patient-contact and non-patient-contact categories. Respiratory therapists, however, had an annual conversion rate of 15.6 percent.
In a study of 56 of 167 North Carolina hospitals reporting data on tuberculin skin test conversions, researchers reported a 5-year mean annual conversion rate for all employees of 1.14 percent (Price et al., 1987). Mean annual conversion rates varied by region of the state (1.80 percent in the east, 0.70 percent in the center, and 0.61 percent in the west). This was consistent with variations in rates for the general population in these sections of the state. The researchers concluded that this association pointed to community rather than workplace origins for new employee infections with M. tuberculosis. In its initial risk assessment for the proposed rule on tuberculosis, OSHA analysts also used these North Carolina data. After critics noted the high prevalence of atypical mycobacteria in the eastern part of North Carolina, the analysts used only the figures for hospitals in the western part of the state to estimate the risk of infection for workers in areas with a moderate prevalence of active tuberculosis. (Other criticisms of the data noted the high nonresponse rate, the limited use of two-step initial testing, and inconsistencies in testing practices. See Chapter 7 for further discussion.)
For its risk assessment, OSHA staff used 1994 Washington State data as a basis for estimating worker risk in low-prevalence areas (62 FR 201). Based on comparisons between hospitals “in zero-TB counties and with no known TB patients” and other hospitals, they estimated that the occupational risk of transmission of M. tuberculosis in Washington State hospitals was 1.5 times higher than the background rate of transmission. For other ways of comparing hospitals, the estimated risk was less. For employees of long-term-care facilities (including nursing homes) and home health care workers, OSAH estimated the risk to be 11 and 2 times the background rate respectively. As noted earlier, OSHA estimated the latter