episodes between 1993 and 1996, two of which involved unrecognized active tuberculosis and two of which were associated with inadequate cleaning and disinfection of bronchoscopes (the latter involving patients only).

Nearly all of the reports summarized above involved inpatient facilities. A few reports have described apparent or documented transmission of tuberculosis in other settings including prisons, jails, funeral homes, and ambulatory care clinics. Most suggest the same general kinds of contributing factors described above. For example, a report on an ambulatory care clinic cites undiagnosed infectious tuberculosis and inadequate engineering controls including insufficient fresh air exchanges in the building and improper ventilation of rooms used for administration of aerosolized drugs (Howell et al., 1989).

Some reports point to unintentional error and inefficiency. For example, according to newspaper reports, a recent outbreak in a Pennsylvania prison involved the improper transfer by the U.S. Immigration and Naturalization Service (INS) of a prisoner with infectious, multidrug-resistant tuberculosis following a “paperwork error” (Hoover, 2000; Lang, 2000). The investigation of this outbreak is, however, incomplete and not yet described in any official report.

Overall, outbreak reports reinforce the picture of implementation presented by surveys and inspections of institutional practices. The reports from the late 1980s and early 1990s underscore the importance of administrative controls, especially respiratory isolation policies, by highlighting the role of undiagnosed infectious tuberculosis and the involvement of particularly susceptible patients as factors in transmission (e.g., those with HIV infection or AIDS). The reports also cite lapses in engineering controls (e.g., lack of isolation rooms and inadequate maintenance).

Several studies suggest that health care workers—including physicians and other professionals—vary greatly in their level of adherence to recommended measures for preventing the transmission of tuberculosis. Chapter 2 has already discussed studies documenting the generally modest rate of compliance of health care workers—including physicians— with recommended treatment for latent tuberculosis infection (Fraser et al., 1994; Blumberg et al., 1996; Ramphal-Naley et al., 1996). Other studies have documented physicians’ incomplete awareness of and adherence to guidelines for treatment of patients with tuberculosis (DeRiemer et al., 1999; Evans et al., 1999). The committee notes that neither the 1994 CDC guidelines nor the 1997 proposed OSHA rule stressed treatment for latent tuberculosis infection. The American Thoracic Society and CDC recently issued guidelines that emphasize the importance of such treatment when indicated (ATS/CDC, 2000b).