Unfortunately, no research has tested individually the effects of respiratory protection on health care workers’ risk of acquiring tuberculosis infection or disease. Some studies do, however, include relevant findings about the mix of measures implemented following hospital outbreaks of tuberculosis in the late 1980s and early 1990s.

Three reports describe hospital responses to well-recognized outbreaks of tuberculosis (two of which involved multidrug-resistant disease) (Wenger et al., 1995; Maloney et al., 1995; Blumberg et al., 1995). In each of the hospitals, the outbreak was ended effectively using variable levels of the tuberculosis control hierarchy. Most important, nosocomial tuberculosis transmission from patient to health care worker was interrupted. Although the hospitals continued to care for substantial numbers of patients with tuberculosis, health care worker exposure incidents and tuberculosis skin test conversions dropped substantially.

Table F-1 summarizes the control measures implemented. Each institution implemented extensive administrative controls, in particular, protocols to promptly identify, isolate, evaluate, and, as appropriate, treat people with signs and symptoms of tuberculosis. Each institution also implemented variable engineering controls, usually some kind of negative-pressure isolation room. Each institution supplied workers with some kind of respiratory protection device.

It is important to note that hospital responses—including the provision of respirators to workers—predated the 1994 CDC guidelines, which specified performance criteria for respiratory protection devices. They also predated NIOSH’s 1995 certification of the N95 respirator, which met the new CDC criteria but was less expensive than previously certified devices. In any case, the respiratory protection measures implemented in these institutions were less stringent than those set forth in the 1994 CDC guidelines, the 1997 proposed OSHA rule on occupational tuberculosis, or the 1998 OSHA