Of the rickettsial agents, Coxiella burnetii probably presents the greatest risk of laboratory infection. The organism is highly infectious and remarkably resistant to drying and environmental conditions (190). The infectious dose of virulent, Phase I organisms in laboratory animals has been calculated to be as small as a single organism. The estimated human ID(25-50) (inhalation) for Q fever is 10 organisms (191). Pike's summary indicates that Q fever is the second most commonly reported laboratory-associated infection, with outbreaks involving 15 or more persons recorded in several institutions (141, 151). A broad range of domestic and wild mammals are natural hosts for Q fever, and may serve as potential sources of infection for laboratory and animal care personnel. Exposure to naturally infected, often asymptomatic sheep, and to their birth products, is a documented hazard to personnel (28, 175). Although rare, C. burnetii is known to cause chronic infections such as endocarditis or granulomatous hepatitis. Genetic analyses (165) as well as structural analysis of the lipopolysaccharides (82) of endocarditis-associated isolates of C. burnetii suggest that specific strains may be associated with endocarditis.

LABORATORY HAZARDS: The necessity of using embryonated eggs or cell culture techniques for the propagation of C. burnetii leads to extensive purification procedures. Exposure to infectious aerosols or parenteral inoculation are the most likely sources of infection to laboratory and animal care personnel (141). The agent may be present in infected arthropods, and in the blood, urine, feces, milk, and tissues of infected animal or human hosts. The placenta of infected sheep may contain as many as 1,000,000,000 organisms per gram of tissue (195) and milk may contain 100,000 organisms per gram.

RECOMMENDED PRECAUTIONS: Biosafety Level 2 practices and facilities are recommended for nonpropagative laboratory procedures including serological examinations and staining of impression smears. Biosafety Level 3 practices and facilities are recommended for activities involving the inoculation, incubation, and harvesting of embryonated eggs or cell cultures, the necropsy of infected animals and the manipulation of infected tissues. Since infected guinea pigs and other rodents may shed the organisms in urine or feces, (151) experimentally infected rodents should be maintained under Animal Biosafety Level 3.

Recommended precautions for facilities using sheep as experimental animals are described by Spinelli (175) and by Bernard (11). An investigational new Phase I Q fever vaccine (IND) is available from the Special Immunizations Program, U.S. Army Medical Research Institute for Infectious Diseases (USAMRIID), Fort Detrick, Maryland. The use of this vaccine should be limited to those at high risk of exposure and who have no demonstrated sensitivity to Q fever antigen. Individuals with valvular heart disease should not work with C. burnetii.

Pike reported 57 cases of laboratory-associated typhus (type not specified), 56 cases of epidemic typhus with 3 deaths, and cases of murine typhus (151). More recently 3 cases of murine typhus were reported from a research facilit (26). Two of these 3 cases were associated with handling of infectious materials on the open bench; the third case resulted from an accidental parenteral inoculation. These 3 cases represented an attack rate of 20% in personnel working with infectious materials.

Rocky Mountain spotted fever is a documented hazard to laboratory personnel. Pike (151) reported 63 laboratory-associated cases, 11 of which were fatal. Oster (144) reported 9 cases occurring over a 6-year period in one laboratory, which were believed to have been acquired as a result of exposure to infectious aerosols.

LABORATORY HAZARDS: Accidental parenteral inoculation and exposure to infectious aerosols are the most likely sources of laboratory-associated infection (89). Successful aerosol transmission of R. rickettsii has been experimentally documented in nonhuman primates (166). Five cases of rickettsial pox recorded by Pike (151) were associated with exposure to bites of infected mites.

Naturally and experimentally infected mammals, their ectoparasites, and their infected tissues are potential sources of human infection. The organisms are relatively unstable under ambient environmental conditions.

RECOMMENDED PRECAUTIONS: Biosafety Level 2 practices and facilities are recommended for nonpropagative laboratory procedures, including serological and fluorescent antibody procedures, and for the staining of impression smears. Biosafety Level 3 practices and facilities are recommended for all other manipulations of known or potentially infectious materials including necropsy of experimentally infected animals and trituration of their tissues, and inoculation, incubation, and harvesting of embryonated eggs or tissue cultures. Animal Biosafety Level 2 practices and facilities are recommended for the holding of experimentally infected mammals other than arthropods.

Level 3 practices and facilities are recommended for animal studies with arthropods naturally or experimentally infected with rickettsial agents of human disease.

Because of the proven value of antibiotic therapy in the early stages of infection, it is essential that laboratories working with rickettsiae have an effective system for reporting febrile illnesses in laboratory personnel, medical evaluation of potential cases and, when indicated, institution of appropriate antibiotic therapy. Vaccines are not currently available for use in humans (see Appendix C).