the pool of most susceptible individuals has been exhausted. Similarly, drops in infection in a workplace could result from the implementation of measures in the community that reduce the number of potential source cases. Some of the studies reviewed below note that new hiring partially refreshed the pool of tuberculin skin test negative workers, and some report that facilities continued to admit substantial numbers of patients with active tuberculosis.
In their review of outbreak reports, Dooley and Tapper (1997) note that the responses to outbreaks have virtually always begun with administrative controls (as recommended and stressed by CDC) to improve prompt identification of people suspicious for active tuberculosis, to make respiratory isolation policies and practices more stringent, and to implement initial treatment regimens that cover the prevalent drug-resistant strains of tuberculosis. The next steps typically involve engineering controls (e.g., the installation, maintenance, and monitoring of negative-pressure isolation rooms). The timing of policies and the specific practices involving respiratory protections appears to be more variable, partly because recommendations about respirators changed several times in the first half of the 1990s.
Below are summarized three of the more complete analyses of associations between facility implementation of tuberculosis control measures and worker risk of infection with M. tuberculosis. Three less complete studies are then briefly described.
Grady Memorial Hospital (Atlanta) After an outbreak of drugsensitive tuberculosis in 1991 and early 1992, Grady Memorial Hospital initiated a number of new tuberculosis control policies and practices during the period from March to July 1992 (Blumberg et al., 1995). The descriptions of the interventions and the subsequent monitoring of skin test conversions and other results are among the most thorough in the literature. The study’s authors conclude that these practices halted the transmission of M. tuberculosis.
Beginning in March 1992, Grady Memorial Hospital implemented a new, more stringent policy of respiratory isolation of patients with known or suspected tuberculosis. Notably, respiratory isolation was required for all patients for whom smears for acid-fast bacilli (AFB) and culture were ordered and for all patients with HIV infection (or risk factors for HIV infection if serology results were unavailable) who had abnormal chest