ABSTRACT
Tuberculosis (TB) infection control in congregate settings such as hospitals, clinics, jails, prisons, hostels, and residential treatment facilities may be the most neglected aspect of global TB control. Although occupational infections with Mycobacterium tuberculosis have been well documented in the United States and Europe over the past century, institutional (including patients, prisoners, students, and other institutional residents) TB transmission in high-prevalence, resource-limited settings has only recently begun to be reported with increasing frequency. From reports limited to hospital workers and outbreaks in other settings, it is difficult to extrapolate the full impact of institutional transmission on the total disease burden at the local, regional, or global level. Yet, there is good reason to believe, especially in high-prevalence settings, that institutional occupants are routinely becoming infected and reinfected at high rates, usually from persons with unsuspected TB. Institutional transmission of TB threatens the health of an already tenuous workforce, especially in areas where HIV/AIDS prevalence is also high. Ultimately, institutional TB transmission undermines efforts to decrease the disease burden in the community. In a rural area of KwaZulu Natal,
South Africa, an analysis of the first 50 cases of extensively resistant Mycobacterium tuberculosis cases showed that 64% had no prior treatment, implying that their infection had been transmitted already highly resistant, and not acquired by erratic treatment as might have been be predicted. Twenty-two percent of these cases had completed therapy and were considered cured, and only 14% had defaulted or failed treatment. Where did this transmission occur? It is always difficult to be sure where transmission occurs, but 56% of these cases had been previously hospitalized and in a rural setting, the hospital would be the most likely place (1). In Lima, Peru, fully 10% of an initial cohort of patients in a community-based multidrugresistant tuberculosis (MDR-TB) treatment program were health-care workers (HCWs), reflecting exposure and infection in local hospitals and clinics (2) (Farmer PE, personal communication. November, 2005). Interns and residents in teaching hospitals in the same city confirm a high risk of nosocomial transmission: an annual infection rate of 17% with 2% active TB (3). In Brazil, the prevalence of TB infection among medical students during the three years of clinical training was 4.6%, 7.8%, and 16.2%, respectively, but in engineering students of the same socioeconomic status the prevalence remained 4.2% to 4.4%, indicating transmission primarily in the hospital setting rather than in the community setting (4). In Malawi, the active TB case rate among 2979 HCWs was 3.2%, or 3200 cases per 100,000 population, compared to 1.8% among primary school teachers (5). At a chest hospital in Estonia, 49 health workers developed TB between 1994 and 1998, with 18 (38%) due to MDR organisms (6). These latter disturbing data recall the U.S. TB resurgence of 1985-1992, where one epidemiologic investigation of 357 genotype-linked cases (one quarter of all MDR-TB cases in the United States at the time) concluded that 96% had likely been transmitted nosocomially, predominantly in just four New York City hospitals (7). When costly interventions reversed the U.S. resurgence, improved case holding as well as improved infection control was credited (8). Similarly, hospital transmission accounted for 68 (88%) of 77 MDRTB cases among HIV patients in a Buenos Aires hospital, and as in the United States, basic infection control measures have greatly reduced spread (9). The problem is not geographically localized, or limited to hospitals or HIV-associated TB. Using extensive contact evaluations and molecular epidemiology, Barnes et al. identified the likely source case and the site of transmission for 79 of 249 TB cases in central Los Angeles, only 27% of which were HIV related. Of the 79 instances of transmission, 55 (70%) were traced to just three homeless shelters (10). Population-based molecular epidemiological studies in New York City and San Francisco have identified a greater proportion (30-40%) of cases due to recent infection than was previously thought possible (11,12). Among recent immigrants to the United States, the rapid decline after arrival in case rates observed among immigrants of all ages has been interpreted as indicating an important role for recent infection and reinfection in the pathogenesis of their disease (13). Although transmission occurs in homes and other noninstitutional
environments, nowhere is transmission more efficient than in congregate settings, and nowhere is there a greater opportunity for interventions to prevention transmission. In a high-prevalence community of South Africa, a molecular epidemiological study estimated that in only 19% of cases was the disease transmitted within the household (14). Given the above data, the importance of controlling TB transmission in institutions cannot be overstated.
