1981 through 1984 (51). The mean time interval was 5 years (calculated ARI = 0.21 percent/year). During the study period, annual conversion rates varied from 0.4 to 1.9 percent, with no apparent secular trend. Tuberculin conversion rates in the institutions’s patients were similar. There was wide variability in the skin testing techniques used, and these figures can be considered only approximate.
Studies in prisons and jails must be considered with the understanding that, among other variables, prisons typically house long-term inmates and jails detain many people for very short periods. The spread of multidrug-resistant tuberculosis among prisoners in New York City and State jails and prisons provoked great concern for the employees of those institutions. For example, in 1988 and 1989, one-quarter of the 205 tuberculosis cases in Nassau County, New York, were associated with a jail (52). Although inmates were screened on admission, there was no screening or infection control program for employees. Statewide, the incidence of active tuberculosis among New York prison inmates increased from 15/100,000/year in 1976 to 139/100,000/year in 1993 (53). Nationally, inmates of correctional facilities contribute just under 2 percent of the tuberculosis case load (48). The age-adjusted case rate for adult inmates is 3.9 times higher than that for the general population.
A system-wide annual tuberculin skin testing program for New York State prison employees was instituted in 1991–1992, and Steenland and coworkers reported on the conversions found at a 1-year follow-up (53). Overall, the conversion rate was 1.9 percent among 24,487 employees. Rates ranged from 1.4 percent in prisons with no known tuberculous inmates to 2.6 percent in prisons with more than the median number of tuberculous prisoners.
Transmission of tuberculous infection from inmates to correctional facility personnel has been documented in several published reports from California penal systems. Two of 11 prison infirmary employees converted their tuberculin skin test after contact with an infected prisoner in 1990–1991 (54). In two other outbreaks, employee tuberculin test conversions occurred in 9 of 319 (2.8 percent) and in 11 of 223 (4.9 percent) employees (55). In all cases, the conversions occurred within 2 years of a previously negative test. In a 1981 outbreak, one employee developed active tuberculosis (56). No information on employee skin test conversions was reported.
A 1994 outbreak of tuberculosis in a Texas prison housing a number of mentally retarded prisoners centered on a classroom used for education of these inmates (57). RFLP analysis demonstrated clustering of the patients. The report does not provide information allowing an assessment