tions, also suggested that the risk of conversions varied considerably by type of hospital, geographic location, job category, and likelihood of contact with high-risk patients. (This review’s discussion of outbreak reports is summarized later in this chapter.)
Another review by Dooley and Tapper (1997) similarly concluded that overall skin test conversion rates for facilities are typically 1 percent or less in nonoutbreak environments, but they emphasized that “overall rates in a facility can mask very high rates in some areas or occupational groups” (p. 368). In studies not specific to outbreak situations, the review authors found mixed results for comparisons of skin test conversion rates for different job categories and different assumed levels of patient contact. Some studies found higher skin test conversion rates for those in jobs with more patient contact (e.g., nursing and respiratory therapy); others did not. The work categorization and other methods and the detail reported in these studies varied considerably (e.g., whether categorizations by patient contact differentiated between contact with patients at high risk of tuberculosis and contact with other patients). The review authors noted that skin test conversion rates in health care workers probably represent a combination of community- and workplace-related transmission of M. tuberculosis. (This review’s discussion of outbreak reports is summarized later in this chapter.)
During the 1980s and 1990s, a number of published articles reported skin test conversion information from state databases, surveys, or studies in one or a few organizations. Most are limited to hospitals.
In a 1987 article entitled “Is the Tuberculosis Screening Program of Hospital Employees Still Required?,” researchers at the University of Washington analyzed skin test conversion data for 1982 to 1984 for 114 hospitals in Washington (Aitken et al., 1987). They put the estimated overall conversion rate for these hospitals at 0.09 percent over the 3-year period (0.03 percent per year), with slightly higher rates for hospitals that admitted tuberculosis patients and slightly lower rates for those that did not. They concluded that the conversion rates in hospitals did not differ significantly from the estimated rate for the state population overall.
A study in a nonoutbreak environment found correlations between positive skin test conversions and the worker’s age, the worker’s race, and the poverty level in the worker’s zip code of residence (Bailey et al., 1995). (Data on tuberculosis case rates were not reported by zip code.) For the period January 1989 through July 1991, the overall rate of skin test conversions was 0.93 percent (0.37 percent annually). After controlling for other variables, the analysts found an association between higher poverty levels and higher rates of positive skin tests and test conversions. Risk