In fall 2001, the United States experienced a bioterrorism attack in which thousands of people, mostly US postal employees, were exposed to Bacillus anthracis (anthrax) spores over the course of several weeks.1 The vectors for distributing anthrax spores in powder form were ordinary envelopes that were routinely processed by the US Postal Service (USPS). In the course of processing these letters, two large USPS processing and distribution centers (PDCs) were widely contaminated. Extensive remediation efforts included multistep surface decontamination/source reduction and facility fumigation with chlorine dioxide (National Research Council [NRC]).2 The fumigation was guided by pre- and postremediation environmental site characterization through culture testing for viable spores. The entire process to remediate and refurbish these two facilities took several years to complete and cost more than $200 million before the buildings returned to normal use.2

Although media reports of anthrax contamination focused on the two heavily contaminated PDCs, other structures sustained more limited anthrax spore contamination, including 21 USPS facilities (3 other PDCs and 18 smaller post offices); private and federal government mail facilities; 7 Capital Hill government buildings (including the Hart Senate Office Building); and commercial buildings and residences. The approach to assessing the degree of contamination, as well as the need and approach to remediation (largely based on the likelihood of past or future potential for aerosolization of spores), varied markedly among the agencies/organizations responsible for these structures. Moreover, countless other governmental, commercial, and private buildings were likely contaminated by cross-contaminated mail, yet few were tested for spores or remediated. As a result, it is quite possible that there are anthrax spores indoors throughout the United States,1 but few public officials would suggest that responsible parties find these sites and clean them. Apart from the incalculable expense, the consensus among knowledgeable experts was that the risk of clinical anthrax infection was too low to warrant this effort,3 a decision that has so far proved to be appropriate since there have been no additional cases associated with these low contamination sites.

In this month's journal, public health officials from New York City (NYC) and Connecticut describe their experiences in 2006 and 2007, respectively, with investigating naturally acquired anthrax cases caused by exposure to spores contaminating imported, untreated West African animal hides used to make drums.4,5 Although only a total of three individuals in the two states were sickened by their exposures, and the indoor environmental contamination appeared to be less extensive than the 2001 contamination of the PDCs, the public health responses to these cases were substantial and costly. Fortunately, many aspects of the 2006 and 2007 responses were informed by the experience of 2001 and the subsequent governmental investment in preparing for the potential of a repeat bioterrorism attack. The joint epidemiologic and law enforcement investigations were done well and were the result of protocol development and joint training that were the direct result of the events of 2001.6 In addition, the recommendations for postexposure prophylaxis and the use of some form of a Unified Incident Command Structure were also based on protocols developed during and subsequent to the attacks of 2001.

Nevertheless, the 2006 and 2007 approaches to environmental sampling (ie, site characterization of contamination) and site remediation were somewhat more ad hoc, partly because of a perceived absence of established protocols to guide the cleanup of naturally occurring anthrax spores in contaminated private facilities. The issues that inform the development of such a protocol are outlined in an NRC report focusing on reopening contaminated public facilities.2 The authors of both articles call for federal leadership to craft protocols and guidelines for local public health officials in the event of future similar incidents of anthrax spores contaminating private buildings. What elements should those protocols contain? Some of the answers to this question may be found in the National Response Team (NRT) Technical Assistance Document for anthrax cleanups, which focuses more on a federal cleanup response to an intentional release incident.7

The basis for decision making on indoor site remediation is characterization of latent environmental hazards through comprehensive, systematic site sampling to assess the extent of site contamination.2 This, in most instances, will allow for a qualitative risk assessment of the possibility of developing either cutaneous or inhalational anthrax. But there are no absolutes to this risk assessment. The 2001 experience informs policy makers of the inherent hazard of relying solely on site characterization to assess the potential for clinical risk. On the one hand, two extensively contaminated PDCs resulted in only nine sick postal employees (three and six postal employees with cutaneous and inhalational anthrax, respectively) despite repeated daily exposures of thousands of postal employees to surface and airborne anthrax spores for weeks.1

On the other hand, cutaneous anthrax developed in a commercial office worker and a postal employee in 2001 in other buildings where environmental sampling revealed only one or no positive cultures, respectively.8,9 Worse still, two individuals died of inhalational anthrax, when neither their workplace (NYC case) nor residences revealed any evidence of anthrax contamination despite very detailed environmental sampling.10,11 Assuming that these latter individuals were exposed to cross-contaminated mail as their source of exposure, and that millions of letters were processed and cross-contaminated in the PDCs between the time the anthrax contaminated letters were first processed and the PDCs were closed, people throughout the United States, and indeed the world, may have received cross-contaminated mail. Yet, no one developed anthrax outside of the three cases likely exposed and made ill though cross-contaminated mail exposure in fall 2001.1

The 2007 Connecticut investigation team concluded that only the drum-maker was at risk for developing inhalational disease. Unfortunately, the degree of site contamination may not give a robust estimate of disease risk, for reasons cited previously. An expert committee stated it clearly: the minimum infectious dose for biological agents such as anthrax is unknown.2 Nevertheless, site characterization needs to be conducted in a systematic fashion, using appropriate and validated sampling methods.7 The approach taken by public healthcare officials and private contractors to assess contamination during and after the 2001 attacks should serve as the basis for establishing a site characterization protocol in a private residence for future anthrax incidents.7,8

Once the characterization results are in hand, officials then must determine the most practical approach to remediation. Federal guidelines must address how best to make this difficult decision, but again there are no absolutes. The approaches taken to remediate buildings contaminated in 2001 varied significantly.2 When there was extensive contamination in USPS PDCs, specifically in Washington, District of Columbia, and Hamilton, New Jersey, the only acceptable approach was to fumigate the entire building under very controlled conditions. Repeated multistep surface cleaning was needed for source reduction before the gassing to minimize the spore burden and to maximize the likelihood of zero growth of B anthracis on culture, the criterion for reoccupancy.2 The size of the facilities (14 and 6.5 million cu ft in Washington, District of Columbia, and Hamilton, New Jersey, respectively) required a systematic grid sampling approach pre- and postfumigation, as was done to some extent in Connecticut in 2007 but not at all in NYC in 2006 where the impacted spaces were small, confined rooms (<12 000 cu ft in total). Other buildings that showed extensive contamination (eg, the Hart building in Washington, District of Columbia, and the American Media, Inc, building in Florida) were also extensively tested and fumigated.2

In contrast, minimally contaminated buildings, as determined by surface swab, wet wipe, and surface vacuum techniques, were cleaned through a four-step process,2 sometimes with localized fumigation,12 with limited postremediation testing, often only where preremediation samples were found to be positive. This approach included postal facilities, commercial buildings, and residences where testing revealed contamination through tracking of spores into a residence from a contaminated workplace. However, most homes of employees who worked in contaminated workplaces were never tested, and thus never cleaned, yet there have been no further cases of anthrax reported that could be linked to the 2001 experience. Clearly, the risk of anthrax from exposure to spores in residences contaminated through tracking of anthrax into the home from work or through cross-contaminated mail is likely to be minimal. But the 2001 Connecticut case illustrates the risk is not zero and could even lead to inhalational anthrax and death.1

In the 2007 Connecticut incident, the case of cutaneous anthrax occurred in the secondarily contaminated home, where testing found growth in 26 percent of samples. It was the potential future inhalational risk posed by numerous contaminated porous surfaces, especially the wall-to-wall carpets, and the secondary cutaneous case that prompted Connecticut health officials in 2007 to fumigate the drum-maker's home. However, the work shed, which was heavily contaminated (100% of samples), underwent only surface decontamination. In contrast, NYC officials concluded in 2006 that surface cleaning was adequate to render the residence and workplace habitable with minimal risk of inhalational anthrax disease in the future, although the index case had inhalational anthrax, which officials ascribed to occupational exposure in his work space. Considerations in the decision-making process included the contamination of the contents of the structures, as well as the quality of the surfaces (less porous surfaces in the workplace than in the home). All in all, the site characterizations were generally comparable in many respects, yet public healthcare officials made different decisions in each jurisdiction. Political and community concerns and public relations had an influence on these decisions, as would be expected.

Development of federal guidance for evaluation, extent of testing, and remediation of anthrax-contaminated private buildings is needed, but we must acknowledge that the guidance would be consensus driven and not based strictly on an extensive scientific foundation, given the experience of 2001 indoor environmental contaminations, gaps in knowledge and the conclusions of the NRC report. Although fumigation is the gold standard of remediation, it, too, may not be absolutely effective or practical because of the scale of contamination, which must be recognized in any risk communication plan. New York City's 2006 communication activities and the NRT report outline what is required to ensure that all stakeholders are regularly informed and updated about real and perceived risks of disease before, during, and after the cleanup process. Not everyone will accept the decisions that are made and a robust risk communication plan is essential to minimize initial and lingering concerns.

A recent case of gastrointestinal anthrax in a woman who participated in a community drum session in New Hampshire, where at least 2 of 66 drums used in the ceremony were contaminated, underscores the environmental risk of using contaminated drums.13 These incidents underscore the importance of the precautionary principle in environmental health14 and the hierarchy of controls in protecting workers. The most important measure in protecting workers and the environment is to use safer materials, wherever possible. Personal protective equipment, if used properly, would offer some degree of protection to the individual working with contaminated hides, but it surely does not prevent environmental contamination, as evidenced in both of these incidents. Safe working practices also might prevent disease and environmental contamination but, too often, even with education about the hazards, people violate their own rules, as in the 2007 incident in Connecticut. The consequences of these failures include self-inflicted infection, bystander disease, public alarm, and a complicated and expensive public health response.

To prevent further incidents as described in this issue, a federal ban on importing untreated raw hides from West Africa and other potential sources, as has been done from Haiti, would be the most protective approach to prevent future incident cases and the public health responses they trigger. Although federally developed guidelines are needed to guide the remediation of an anthrax-contaminated private facility, it is not a substitute for a regulatory measure to prevent the importation of contaminated hides or other products. As the saying goes, an ounce of prevention is worth a pound of cure (or remediation).

REFERENCES