An Outbreak of Multidrug-Resistant Tuberculosis among Hospitalized Patients with the Acquired Immunodeficiency Syndrome

TLDR

Since 1990, several clusters of multidrug‑resistant tuberculosis have been identified among hospitalized patients with AIDS. The authors aimed to alert clinicians to the risk of nosocomial MDR TB transmission and to emphasize strict adherence to infection‑control measures. They conducted a retrospective case‑control study comparing exposures of 18 MDR TB patients with 30 controls with drug‑susceptible TB and 14 MDR TB patients during the six months before diagnosis with 44 matched controls, using hospital ward data and exposure periods. The analysis showed that MDR TB patients were seven times more likely to have been hospitalized during exposure periods and 52 times more likely to share a ward with an infectious MDR TB patient, with 15 of 16 isolates sharing identical RFLP patterns and only one of 16 rooms having negative‑pressure ventilation, demonstrating rapid nosocomial transmission among AIDS patients.

Abstract

Since 1990 several clusters of multidrug-resistant tuberculosis have been identified among hospitalized patients with the acquired immunodeficiency syndrome (AIDS). We investigated one such cluster in a voluntary hospital in New York.We compared exposures among 18 patients with AIDS in whom tuberculosis resistant to isoniazid and streptomycin was diagnosed from January 1989 through April 1990 (the case patients) with exposures among 30 control patients who had AIDS and tuberculosis susceptible to isoniazid, streptomycin, or both. We also compared exposures among the 14 case patients hospitalized during the six months before the diagnosis of tuberculosis (the exposure period) with those among 44 control patients with AIDS matched for duration of hospitalization. Mycobacterium tuberculosis isolates were typed with analysis of restriction-fragment-length polymorphism (RFLP).Case patients with drug-resistant tuberculosis were significantly more likely than controls with drug-susceptible tuberculosis to have been hospitalized during their exposure periods (14 of 18 vs. 10 of 30) (odds ratio, 7.0; 95 percent confidence interval, 1.6 to 36; P = 0.006). Case patients hospitalized during their exposure periods were significantly more likely to have been hospitalized on the same ward as a patient with infectious drug-resistant tuberculosis than were either controls with drug-susceptible tuberculosis hospitalized during their exposure periods or controls matched for duration of hospitalization (13 of 14 vs. 2 of 10 and 23 of 44) (odds ratio, 52; 95 percent confidence interval, 3.1 to 2474; P less than 0.001; and odds ratio, infinity; 95 percent confidence interval, 2.4 to infinity; P = 0.005, respectively). Among those hospitalized on the same ward, the rooms of case patients were closer to that of the nearest patient with infectious tuberculosis than were the rooms of controls matched for duration of hospitalization. M. tuberculosis isolates from 15 of 16 case patients had identical patterns on RFLP analysis. Of 16 patients' rooms tested with air-flow studies, only 1 had the recommended negative-pressure ventilation.Multidrug-resistant tuberculosis is readily transmitted among hospitalized patients with AIDS. Physicians must be alert to this danger and must enforce adherence to the measures recommended to prevent nosocomial transmission of tuberculosis.