CDC anthrax-related field investigations from 1950 to 2001 were identified from several sources. First, the new database of historical Epi-Aid documents (1950–1999) was searched to retrieve all documents in which “anthrax” or “anthracis” appeared either as an assigned keyword or as a text string in a full-text search. Epi-Aid documents related to anthrax investigations in 2000 and 2001 were identified manually in an EIS administrative database. These searches identified a variety of types of documents, including initial requests for epidemiologic assistance, interim progress reports, final reports, and memoranda.

To identify published reports on these Epi-Aid investigations, we searched indexes to the Morbidity and Mortality Weekly Report (MMWR) for anthrax-related reports for the years 1961–2001. The individual issues of MMWR and its predecessor (Weekly Morbidity Report) were searched manually for the years (1950–1960) for which no index exists. To identify published reports on anthrax-related Epi-Aid investigations, we searched Medline for the years 1966–2001 and Index Medicus for 1950–1965. The names of the lead investigators from the Epi-Aids were used as keywords.

Additional CDC anthrax-related field investigations were identified by two coauthors (PB and AK) who were personally involved in most anthrax investigations conducted by the agency since the 1950s. References describing these additional investigations were located in the MMWR and in published medical articles. To limit this report to a description of CDC’s institutional experience, rather than a broader review of publications on anthrax investigations, we excluded (a) anthrax case reports published in the MMWR but unrelated to a CDC field investigation and (b) published reports on anthrax by investigators not affiliated with CDC.

From the unpublished Epi-Aid documents and published reports for each investigation, we abstracted the following information: year, location, number of human and animal cases, clinical form of the disease, occupational or other exposures for human patients, environmental sampling methods and data, and study recommendations.

A total of 49 relevant field investigations (Table 2) were included in this report: 42 Epi-Aids and 7 other investigations. Detailed reports and MMWR published summaries were available for 39 (93%) of the 42 Epi-Aid field investigations. For three agriculture-related investigations (Epi-Aids 1963-2, 1959-38, 1957-17), only the initial invitation for epidemiologic assistance was available for review.

Of these 49 field investigations, 41 (84%) involved human or animal infections with B. anthracis, 2 were evaluations of decontamination of B. anthracis–contaminated textile mills (1967 and 1972), and 1 was an investigation of bioterrorism threats involving anthrax (1998). In the other 5 investigations, B. anthracis was not found to be the causative organism, despite initial suspicion. Because these investigations include only anthrax cases for which CDC’s assistance was requested, they represent only a small proportion of the total number of U.S. cases reported during this period (Table 1).

Most of the investigations (41/49, 84%) were conducted from 1950 to 1980; only 8 anthrax-related investigations were conducted by CDC from 1980 until the October 2001 bioterrorism events. This trend mirrors the decline in reported U.S. anthrax cases in the latter half of the 20th century (10) (Table 1).

Field investigation reports usually contain public health recommendations; many of these are appropriate for future anthrax epidemics or exposures. For infections associated with textile mills, a 1974 report stated that “decontamination of the primary source of B. anthracis is not generally held to be practical” (Epi-Aid 1974-77). The reports on textile mill investigations recommended anthrax vaccine with annually scheduled booster inoculations for mill workers at risk; use of personal protective equipment including specific work clothing and respirators, shower facilities, and separate lockers for work and street clothing; physical separation of raw and finished materials to prevent cross-contamination; design of work areas for easy cleaning; and air-exhaust systems designed to prevent the spread of spores. One report recommended that mill employees be “thoroughly indoctrinated” on the cause, nature, and control of anthrax (Epi-Aid 1953-14). In 1999, following multiple bioterrorist threats (38) (Epi-Aid 1999-25), antibiotic prophylaxis was recommended in cases with known or credible risk for direct exposure. For persons with suspected exposure to aerosolized spores, recommendations included isolating exposed clothing in a plastic bag, showering with copious amounts of soap and water, and washing all possibly contaminated materials with a 1:10 bleach dilution (38).

For infections associated with farms and livestock, reports recommended vaccination of animals at risk, better education of farm workers on anthrax diagnosis and control, thorough destruction by burning of infected animals, prevention of infected livestock from reaching the market, improved supervision of slaughter and meat inspection, and, in some situations, farm quarantine. After the 1974 Texas epizootic (Epi-Aid 1975-06), anthrax vaccine was tested in dairy cattle to assure that the vaccine had no adverse effect on milk safety (6).

Investigations of B. anthracis–contaminated saddle pads (1974), Haitian handicrafts (1974), and imported yarn (1976) led to Consumer Product Safety Commission recommendations for destroying those products (7,39,45). In 1975, cutaneous anthrax developed in a New Jersey gelatin manufacturing plant worker after his exposure to contaminated dry cattle bones; the Occupational Safety and Health Administration levied fines for workplace safety violations (41).

The Haitian investigation also led to a federal ban on importing Haitian goatskin products. A review of such handicrafts collected at U.S. quarantine stations in 1980–1981 found that items continued to be contaminated with B. anthracis (47). Recommendations to the Haitian Ministry of Health included providing incentives for reporting diseased animals, improving laboratory diagnostic capacity, increasing anthrax vaccination levels among livestock, educating livestock owners about the benefits of anthrax control, and improving the tanning procedures for goatskin drum heads (Epi-Aid 1974-96).

In this report we review what has been learned from >40 epidemiologic field investigations of confirmed or suspected anthrax outbreaks in humans or animals during the last 50 years. In the 2001 bioterrorism response, investigators evaluated suspected anthrax cases by using clinical and laboratory diagnostic methods, such as chest radiographs, cultures, and serologic assays, that had been developed and refined during earlier investigations of inhalational and cutaneous anthrax in textile mill workers. In addition, histopathologic and immunohistochemical testing proved essential for diagnosing anthrax in persons who had been placed on antibiotics early and whose cultures were thus negative. Nasal swabs, as used in the 1953 textile mill investigation, are currently considered an unevaluated adjunct to environmental sampling for defining exposed populations in bioterrorism investigations (1,60). Nasal swabs were used in the 2001 investigation for defining the aerosol spread of B. anthracis spores in the Hart Senate Office Building and some other settings.

In the recent investigation, an anti-protective antigen, enzyme-linked immunosorbent assay (61) was used to confirm B. anthracis infection in several cases. Development of this assay was the culmination of decades of laboratory experience and research associated with past field investigations of anthrax.

Asymptomatic infection was documented in one serologic survey (33) conducted several months after an inhalational anthrax outbreak; however, in past and current investigations, the role of asymptomatic infection in providing protection is unclear. Human-to-human spread was not evident in any of the investigations reviewed.

Investigation into a series of anthrax-related threats and hoaxes in 1998 (Epi-Aid 1999-25) also helped lay the groundwork for the recent response. In that investigation, guidelines for risk assessment and postexposure antibiotic prophylaxis were developed, and coordination with first responders and law enforcement was emphasized (38). The investigation also led to revised immunization recommendations (5), which discuss the use of vaccine for postexposure prophylaxis.

In response to the bioterrorism events of 2001, additional guidelines were published on investigating and responding to B. anthracis exposures. These address clinical testing, use of antibiotic prophylaxis, closing of potentially contaminated buildings, and postexposure treatment options (1,62,63). Current recommendations for the use of anthrax vaccine are based in large part on a field trial conducted in 1962 (2,5). During the 2001 response, vaccination recommendations were expanded to at-risk populations; the 1962 vaccine efficacy study forms part of the justification for considering the vaccine for postexposure prophylaxis. Currently, the Advisory Committee on Immunization Practices recommends that vaccine be used in combination with antibiotics (ciprofloxacin, doxycycline, or penicillin) following a B. anthracis bioterrorism exposure, if vaccine is available (5). Vaccination is a critical component of the nation’s preparedness and response activities for B. anthracis bioterrorism.

In past field investigations, the primary risk factor for human cutaneous anthrax has been direct physical contact with infected animals or commercial products containing B. anthracis spores. Ranchers, butchers, and veterinarians were at risk for such contact when working with infected animals. All the commercial products causing human infection were of animal origin; most were made from imported goat skin or hair.

For inhalational anthrax, the main risk factor was exposure to aerosolized spores, especially in or near a textile mill that processes goat hair. While it is unclear why some workers become infected while others in the same dusty environment do not, several factors may increase the likelihood for infection. First, direct work with unprocessed goat hair may create a heavier exposure to B. anthracis spores. Second, a weakened immune system may increase a person’s susceptibility to infection (64). Two of the patients with inhalational anthrax probably had chronic pulmonary disease. In the 1957 investigation, sarcoidosis was present (31). In the 1966 investigation of a metal shop worker (Epi-Aid 1967-43), investigators noted the worker’s “chronic cigarette cough” and suggested that his alcoholism, diabetes, and pancreatitis might have made him more susceptible than his healthy coworkers.

Over the past 50 years, a series of recommendations have focused mainly on preventing occupationally acquired anthrax, especially in textile mills and agricultural settings. For example, in 1962, anthrax vaccine was recommended for persons who handle imported hair, wool, hides, or bonemeal (2). More recently, it was recommended that veterinarians obtain diagnostic specimens but not perform necropsies on animals suspected to have died from anthrax (36). The National Institute for Occupational Safety and Health has been actively involved in many recent anthrax-related investigations (15,65).

Some documents mentioned insects as possible vectors in the spread of B. anthracis. While mechanical spread of B. anthracis organisms by stable flies has been demonstrated in guinea pigs (66), the importance of insects as vectors in epizootics has not been determined. One hypothesis suggests that insect bites might allow superficial organisms an effective access point for intradermal infection. Insects, particularly horseflies, were explicitly mentioned in 12 investigations for their possible role in transmission; however, no evidence exists that biting flies contribute to transmission of disease from animals to humans.

Past methods for decontaminating buildings relied upon formaldehyde gas, now known to be carcinogenic. The recent decontamination of B. anthracis–contaminated buildings was accomplished with chlorine dioxide gas, by using the methods developed for decontaminating textile mill buildings. Pre- and posttreatment environmental sampling strategies developed in several of the earlier field investigations, including the systematic use of surface swabs and spore strips, were also used in the response to recent events. In these events, the wide dispersion from envelopes of small airborne particles containing spores led to higher than expected levels of cross-contamination, making decontamination more difficult (65).

Several limitations should be considered in interpreting the results of this review. CDC conducts field investigations only when invited by a state health department or ministry of health. Anthrax cases that did not actively involve CDC staff, such as those investigated solely by state or local health departments, were excluded; therefore, this is not a complete report of U.S. anthrax case investigations. However, CDC staff have consulted at least by telephone on almost every case of human anthrax reported in the United States since the 1950s (A. Kaufmann, pers. comm.). A manuscript reviewing the characteristics of all anthrax cases reported in the United States since 1955 is in preparation (D. Ashford, pers. comm.). Second, this review examines CDC’s experience with field investigations involving anthrax; laboratory-based anthrax research was not included unless it was related to a field investigation. Third, final laboratory results were not available for some field investigations.

Much useful knowledge, ranging from the diagnosis of anthrax to the use of vaccine to protect populations, has been gained from these past investigations. However, many questions remain. Further research is needed to determine the lowest infectious dose, define what constitutes a true exposure for which antibiotic prophylaxis is warranted (especially in light of possible drug side effects), and determine whether spores delivered in an envelope create a residual risk after the primary contamination event. Other areas in which more research is needed include developing better rapid environmental testing methods (67), identifying optimal decontamination methods for a variety of contaminated settings, assessing B. anthracis spore background rates in selected settings, and determining the level of risk associated with a low degree of exposure to aerosols containing B. anthracis.

During the past 50 years, the scientific knowledge acquired in these field investigations has greatly improved the nation’s ability to respond to anthrax outbreaks. New and unique challenges have been raised by the recent intentional release of B. anthracis. Further efforts to improve knowledge about anthrax, both in its natural setting and in the context of bioterrorism, are urgently needed.