BIOSECURITY AND BIOTERRORISM: BIODEFENSE STRATEGY, PRACTICE, AND SCIENCE Volume 3, Number 3, 2005© Mary Ann Liebert, Inc.

Postal Workers’ Perspectives on Communication During the Anthrax Attack

SANDRA CROUSE QUINN, TAMMY THOMAS, and CAROL McALLISTER

In 2001, the nation experienced its first bioterrorism attack, in the form of anthrax sent through theU.S. Postal Service, and public health professionals were challenged to communicate with a criticalaudience, U.S. postal workers. Postal workers, the first cohort to receive public health messages dur-ing a bioterrorist crisis, offer a crucial viewpoint that can be used in the development of best prac-tices in crisis and emergency risk communication. This article reports results of qualitative inter-views and focus groups with 65 postal workers employed at three facilities: Trenton, New Jersey;New York City; and Washington, DC. The social context and changing messages were among thefactors that damaged trust between postal workers and public health professionals. Lessons learnedfrom this attack contribute to the growing body of knowledge available to guide communications ex-perts and public health professionals charged with crisis and emergency risk communication withthe public.

IN 2001, THE NATION EXPERIENCED a bioterrorist attack,in the form of anthrax sent through the U.S. Postal Ser-

vice (USPS). By November 2001, 22 anthrax cases, in-cluding 11 cases of inhalational anthrax, were confirmed,with most cases linked to letters mailed in Trenton, NewJersey.1 Nine cases were postal employees, including sixdiagnosed with inhalational anthrax.1 The Hart SenateOffice Building, where anthrax was released from a letterto Senator Tom Daschle, closed October 17, 2001, andthe following day the Trenton Postal Processing and Dis-tribution Center, where two postal workers were diag-nosed with inhalational anthrax, closed.2,3 Between Octo-ber 19 and October 21, four workers at the BrentwoodPostal Processing and Distribution Center in the Districtof Columbia were hospitalized with inhalational anthrax;two of these workers—both African Americans—died.The building was closed on October 21.4 Anthrax wasfound in the Morgan Central Postal Facility in New YorkCity, but that facility remained open. Ultimately, 8,424postal workers received a 60-day course of antibiotics.5

Recognizing the new communication challenges, theCenters for Disease Control and Prevention (CDC) de-veloped a model for crisis and emergency risk communi-cation (CERC), defined as:

an effort by experts to provide information to allow anindividual, stakeholder, or an entire community to makethe best possible decisions about their well-being withinnearly impossible time constraints and help people ulti-mately to accept the imperfect nature of choices duringthe crisis.6(p6)

Crisis and emergency risk communication reflects thechallenges of such communication in key ways: (1) deci-sions must occur in a compressed time frame; (2) the decision may be irreversible; (3) the outcome may be un-certain; and (4) the information necessary for that deci-sion may be incomplete or uncertain.6(p6)

Although there is an extensive body of literature onrisk communication, including examinations of the rela-tionship among trust, credibility, and risk perception, lit-

Sandra Crouse Quinn, PhD, is Associate Professor; Tammy Thomas, MSW, MPH, is Project Director; and Carol McAllister,PhD, is Assistant Professor; all are in the Department of Behavioral and Community Health Sciences, Graduate School of PublicHealth, University of Pittsburgh, Pittsburgh, Pennsylvania.

207

tle of it focuses on an actual bioterrorism event. How-ever, the growing literature on crisis and emergency riskcommunication concludes that trust is critical to its suc-cess.6–11 Shore defines trust as “an unwritten agreementbetween two or more parties for each party to perform aset of agreed-upon activities, without fear of change fromeither party.” 8(p13)

The literature suggests that trust is comprised of caringand empathy, competence and expertise, dedication andcommitment, and honesty and openness. Some authorsalso include fiduciary responsibility, absence of bias,predictability, and fairness.7,12–14 Trust can diminishwhen there is disagreement among experts, lack of coor-dination among organizations, unwillingness to acknowl-edge risks, unwillingness to disclose information, per-ceived irresponsibility in managing risk, or insensitivityon the part of authorities to the public’s need for dialogueand participation.7

Social trust in institutions may be especially importantduring unfamiliar events; in fact, Frewer asserts that“trust is likely to be particularly important under circum-stances where people feel they have very little personalcontrol over their personal exposure to potential haz-ards.”15(p393) Credibility also may be particularly relevantto successful communication during a bioterrorist event.Trumbo and McComas found that the higher the credibil-ity of the health department, the lower the perceived riskby the public.16

Several other studies have examined the anthrax at-tack, and two qualitative studies with postal workersfrom the Brentwood facility and congressional staff fromCapitol Hill identified trust as a critical issue.17,18 Thesestudies found that postal workers reported getting muchof their information from the media and from co-work-ers, and they saw the information they received as un-clear and not timely. The researchers also found that in-accurate information and lack of consistent messagescontributed to erosion of trust in public health agencies.Furthermore, perceived differences in treatment becauseof race and class damaged trust significantly, withAfrican American postal workers more likely to believethey were being experimented on in ways similar to theTuskegee syphilis study. While there was some recogni-tion that public health agencies did not have much infor-mation on anthrax, participants also perceived a lack ofexpertise that hampered communication.17,18

The GAO study of the Wallingford (Connecticut) fa-cility also revealed that lack of timely and understand-able communication affected trust; in fact, a report fromthe Occupational Safety and Health Administration to theGAO stated, “Failure to effectively communicate issues,which can have an effect on a worker’s health and safety,can lead to fear and mistrust.”19(p6) In contrast, Mullin20

argues that Mayor Giuliani’s success with communica-

tion during the anthrax attack was based on his ability todemonstrate empathy and credibility with his team of ex-pert department heads.

Two studies suggest that the early failure to identifypostal workers as an at-risk audience created communi-cation challenges during the attack.17,21 A case studyabout communication in New Jersey describes the needfor public health agencies to identify audiences who maynot be obviously at risk. The study suggested that failureto triage audiences, thereby identifying audiences whoperceived themselves to be at risk because of occupationor other characteristics, contributed to a lack of commu-nication with postal workers at the Monmouth facility.21

This study also found that the inability or unwillingnessof agencies to address confusion and contradictionsabout nasal swabbing damaged communication efforts.

Clearly, there is much to learn about crisis and emer-gency risk communication. To date, the present study isthe first with postal workers from the three major sites af-fected by the attack. This study examines the reactions ofpostal workers to the anthrax attack and provides lessonsin crisis and emergency risk communication for futurebioterrorist events.

METHODS

This study used a qualitative case study design to ex-plore the impact of the 2001 anthrax attack at the threepostal facilities most seriously affected by this event: theBrentwood Postal Processing and Distribution Center,the Trenton Postal Processing and Distribution Center,and the Morgan Facility in New York. Given the casestudy format and our use of qualitative methods, ourfindings cannot be generalized in any statistical sense tothe whole population of postal workers either during the2001 attack or prospectively in terms of future bioterror-ism events.

However, our purpose was quite different. We soughtan in-depth examination of experiences of postal workersin order to develop our understanding of worker reac-tions to threats of bioterrorism and to the intervention ofpublic health authorities. Our goal was to gain insightinto the general problem of crisis and emergency riskcommunication, rather than to extrapolate study findingsto the diverse population of postal workers.22

Our research project required that trust be developedbetween the research team and participants. Initial trustwas established primarily through discussions with unionleaders and worker organizations at each facility, allow-ing researchers to explain the purpose and plan of thestudy and giving postal workers an opportunity to raiseconcerns and questions. Following this process of rela-tionship building, nine key informant interviews were

QUINN ET AL.208

completed in the first 6 months of the study. Some keyinformants were identified from newspaper coverage,while others were recruited through “snowball” samplingwith initial interviewees, who identified other postalworkers who had played a significant leadership role dur-ing the crisis (see Table 1 for description of key partici-pants).

Interviews with key informants employed an interviewguide developed from risk communication literature andmedia coverage of the attack. Following best practices inqualitative research, new issues that were raised by earlyinformants were incorporated into subsequent inter-views. Discussion topics included the role of the key in-formant during the crisis, postal workers’ experiencesduring the crisis, risk perceptions, how and where infor-mation was obtained, the perceived trustworthiness ofsources, preferences for public health response, appropri-ate spokespeople, and the anthrax vaccine. The majorityof the interviews were completed in person.

Interviews with key informants were followed by fo-cus groups and individual interviews with rank-and-fileworkers. These combined methods allowed workers toselect the interview format most comfortable for them,and researchers were able to explore key questionsthrough workers’ recounting personal experiences in in-dividual interviews as well as through interactive discus-

sions in focus groups. Discussion guides for both meth-ods delved deeper into the major themes that emerged inkey informant interviews.

One focus group was conducted in Washington, two inNew York, and three in New Jersey. Focus groups wereconducted by an experienced facilitator in settings identi-fied by workers as being safe and neutral: a union office,a church, and a public library. The project coordinatorconducted telephone interviews with 20 postal workers.Recruitment strategies included flyers posted in postalfacilities and distributed by union representatives, adver-tisements in union newsletters and on union email list-servs, word-of-mouth from co-workers, and personal in-vitations by union leaders. Although participants wereself-selected from the total workforce at the facilities,most of the 65 participants (including key informants)are African American, they represent numerous nonman-agement craft positions, and they had considerable tenurewithin the USPS system (see Table 1).

All individual and focus group interviews were tapedand transcribed. Two researchers reviewed the transcriptsand developed basic categories or codes that describedthe key topics of discussion and issues of concern. In or-der to identify themes, each of these coded categorieswas then further examined, compared across types of in-terview (key informant, individual worker, focus group),

COMMUNICATION DURING THE ANTHRAX ATTACK 209

TABLE 1. KEY INFORMANT, INDIVIDUAL INTERVIEW, AND FOCUS GROUP PARTICIPANTS

Individual interviews andParticipants Key informants focus groups

CraftClerk 3 33Maintenance 2 3Mail Handler 2 19Technician 1 0Motor Vehicles 1 1

Tenure with USPSRange 15–35 years 6–38 yearsMean 23 years 21 years

AgeRange 40–63 years 32–69 yearsMean 53 years 49 years

RaceAfrican American 7 38White 2 13Hispanic 2African American–Hispanic 3

GenderFemale 0 28Male 9 28

and discussed with a third researcher. Ideally, qualitativedata analysis is continuous, occurs during the same timeframe as data collection, and is iterative in nature. In thisstudy, insights emerging from early stages of analysiswere important for the refinement of questions and topicsfor subsequent interviews or focus groups and pointedthe way toward overarching constructs that were furtherdeveloped and refined through the analysis of the com-plete set of interviews.

Three themes are explored related to the problem of ef-fective crisis and emergency risk communication:

• the necessity of adapting crisis and emergency riskcommunication strategies to the social and politicalcontext in which risk communication occurs;

• the problems of uncertainty, changing information, andmixed messages in rapidly changing situations such asa bioterrorist attack; and

• the essential qualities that characterize effective crisisand emergency risk communication—in particular,trust, credibility, caring, empathy, openness, and hon-esty—and possible means of demonstrating such qual-ities in a crisis.

RESULTS

While investigations and interventions differed acrosssites, participants universally discussed the lack of ade-quate communication through all phases of the crisis.They reported that the early responses of public healthprofessionals reflected the fact that they did not fully ap-preciate the social and political context in which this cri-sis occurred. Additional constraints placed by the Na-tional Response Plan and agency restrictions influencedpublic health professionals’ response, which was inter-preted negatively by workers, thus beginning a down-ward spiral that led many to a total distrust of the publichealth system. These early challenges will be discussed,leading to the larger discussion on the erosion of trust.

Understanding Communication in the Social and Political Context

While many participants followed the investigation ofthe attack from its inception, a larger number became in-terested only after discovering that letters were the con-duit for anthrax. This realization led to concern abouttheir own risk. Participants reported that early communi-cations from USPS assured them that there was little orno risk. They recalled being told that it was “highly un-likely” that they were at risk and that they had “morechance of being hit by a car.”

However, the postal workers knew that letters pro-cessed through high-speed machines are prone to dam-

age, so they began to question experts’ assumptions andseek their own information. They reported that the pre-ventive measures that had been suggested, such as handwashing and wearing gloves and protective masks,seemed inadequate. Participants noted the procedure forcleaning the mail processing machines, which entailedblowing the contents of the machine directly into the fa-cility. Knowing that machines that processed contami-nated letters had been cleaned by this method, workerssuspected anthrax contamination throughout their work-site. One worker described it this way: “Wash yourhands. This doesn’t make any sense to me. We are atgreat risk here. We used to blow the machines out.”

While many participants said they initially trusted theinformation received from health officials “who assuredus not to worry, that things were fine,” as events unfoldedand reassuring statements proved inaccurate, confidencein professional advice waned. Participants reported thatas they became increasingly aware of their risk, they alsowanted assurance of their safety; with information fromofficial sources in short supply, they pursued othersources.

Participants viewed the media as a reputable source ofinformation. They continued to use multiple media, in-cluding newspapers, television, and the Internet, through-out the crisis. While at work, some participants reportedlearning more about their facility either directly from themedia or through personal networks reporting on mediacoverage: “I had people calling me from all over theplace giving me more information than we got at work.”However, participants sought information from multiplesources to confirm facts and determine the credibility ofthe information. One worker described the process thisway: “Listen to stuff and turn from channel to channel tosee if they are all saying the same thing . . . that is a goodthing. You have to draw your own conclusions.”

Once the USPS acknowledged that the facilities hadprocessed letters containing anthrax, the advice on riskreduction was more forthcoming but still seemed insuffi-cient to participants. Word of mouth from peers and per-sonal networks, used from the start to share messages, remained a trusted means of sharing information. Partici-pants across sites described a process whereby “youcalled somebody, they called somebody, they calledsomebody.”

Workers consulted their personal physicians to assistthem in making health and safety decisions. To varyingdegrees, union locals took the initiative to provide infor-mation to their constituents. Although many participantsused the Internet for information throughout the crisis,information on the Internet sometimes contradicted in-formation received from official sources. For some, thisstirred mistrust: “When I looked on the Internet, that’swhen I found out they lied to us all the way about an-

QUINN ET AL.210

thrax.” Participants without Internet access often re-ceived web-based information from co-workers or familymembers.

Communication during the attack was shaped by thecontext in which it occurred. A long-standing contentiousrelationship between USPS management and laborunions influenced the reaction of postal workers and cre-ated barriers for public health professionals who were at-tempting to build trusting relationships. Participantsviewed health officials as partners with USPS manage-ment: “I’m still angered over the fact that they are in ca-hoots with the postal service. So did we trust them? No.”For some participants, this association alone was a causefor distrust. For others, they understood that public healthagencies, including CDC, had to work with the USPS, aquasi-government agency. However, they wanted to seesome demonstration of CDC’s commitment to protect thehealth of workers. Others suggested that the CDC neededto be independent of postal management: “I’d like to hearfrom them without any postal [management] peoplethere.” Another said, “Give the true impression that CDCis completely independent of the postal service in dealingwith employees.” Some felt the USPS dictated what theCDC was able to tell employees: “Even if managementdoesn’t want them to tell us, they need to tell us.”

Many participants reported no previous knowledge ofthe CDC, and, consequently, they had no understandingof CDC’s role and authority in the crisis response. Al-though CDC’s role was constrained by the multiagencyinvolvement, many participants regarded CDC as beingthe highest authority. The comments in Figure 1 reflectparticipants’ expectations. According to participants,their assumptions and unfulfilled expectations contrib-uted to anger and declining trust in the CDC. The CDCreceived the majority, but not all, of the criticism. Localand state health departments also were criticized for theirperceived lack of authentic concern and response.

Mixed Messages

In the rapidly unfolding event, participants felt that itwas difficult to make decisions about their health be-cause “things were bouncing around so quickly.” As thescientific investigations yielded new information, publichealth authorities communicated new messages in an at-tempt to keep workers and the public informed. Unfortu-nately, many participants interpreted changing messagesas a lack of honesty and an attempt to hide information.As one worker summarized, “Maybe they were hidingsomething. They weren’t really being as honest. Theywould give us a little, not the whole thing.” This suspi-cion extended to those instances in which public healthprofessionals and the USPS management provided dif-ferent information: “Somebody had something to hide. . . we felt we weren’t getting the truth.” According toparticipants, these mixed messages were troubling, espe-cially in the context of the frightening event. One partici-pant stated:

Today, they say it can’t kill you. Then tomorrow, theysay . . . you get enough in your system, then it can killyou but either way you are scared on the first day. Youare still going to be scared on the second day. You justhope you don’t die.

Many participants reported that contradictions betweenverbal reassurances and actual behavior made them angry.When teams dressed in protective clothing conducted envi-ronmental testing, participants responded, “When theybrought in the CDC people or these guys in their little spacesuits, they were still continuously telling their employees, ifyou see the people in there with the spacesuits on, don’t be-come alarmed. They are just testing.” In response, one par-ticipant said, “My first thought was, if the building is sosafe, why do those men need to dress that way when theylet us work all day?” Another worker responded,

COMMUNICATION DURING THE ANTHRAX ATTACK 211

➢ “I think they should have stepped in, in the beginning, and forced the US Postal Service.”➢ “I think they should have had the authority to shut Brentwood down. I think CDC can

quarantine areas, don’t they?”➢ “CDC should have had the same authority to go out there and seize the building.”➢ “They have the knowledge of knowing what anthrax was; the postal officials didn’t. They

could have said we want this building shut down.”➢ “. . . if the postal service kept telling us, we’re getting our information from the CDC, then

I think they should have stepped forward and been the front people on this. Been at thenews conferences, been at our meetings, visited the post office, tried to talk to our doctors.”

➢ “CDC has the authority and the power to regulate any agency where they are in violation ofsafety and health, and they have the power to bring in other agencies to make sure that theproper testing is done. They didn’t do that.”

FIGURE 1. POSTAL WORKERS’ EXPECTATION OF PUBLIC HEALTH AUTHORITIES.

I felt angry and betrayed because there was a point wherewe would be at work and you would see people or teamscome in hazmat uniforms, and we worked and suspectedthe danger. What they should have done was evacuatethe building and conduct the research. . . . You come inthere in a hazmat uniform walking right by me and I amworking. There is a problem there. I am doomed.

Participants clearly identified a crucial need for timelyand accurate information. As one union leader reflected,“Employees need to know as much information as possi-ble so that they can make informed decisions about theirhealth.” However, the timeliness of information on riskwas problematic. One worker commented, “The informa-tion was going to save your life. When that informationwas needed, everybody was absent. These [anthrax]meetings didn’t get in place until the disaster was likeleveled out. I mean these meetings should have happenedfrom day one. . . . ”

Across the sites, participants agreed they were not keptwell informed as events unfolded, resulting in confusionand much concern about personal safety. One workercommented, “I understand that sometimes they don’twant to create hysteria, but you have to let the publicknow immediately what to look out for, symptoms. . . . ”They wanted public health officials to intervene withsound information to allow workers to make decisionsabout their risk. As one participant stated, having infor-mation allowed him some autonomy: “I don’t want youto put my situation in your hands. Tell me what I amdealing with. Let me make that decision and not you.”

Credibility, Openness, Honesty, Empathy,Caring, and Trust

CredibilityTrettin and Mushan define credibility as being believ-

able, trustworthy, and reliable.13(p411) While participantswanted public health professionals to demonstrate knowl-edge about anthrax, they reported being frustrated becauselittle information was available about manmade anthrax.Some participants described public health officials as “notvery knowledgeable” and said, “They couldn’t give anyconcrete answers on anything.” A common refrain was: “Iam asking you because you are the expert and you have noanswers to give me.” Participants expected public healthprofessionals to be prepared: “But public health shouldknow something about something before it happens.”“Why didn’t CDC have more information, because theyhave been studying this for over 20 years?”

The distinction between naturally occurring and new“weapon”-grade anthrax was not clear to all participants.However, others understood that much of what profes-sionals knew was about naturally occurring anthrax, not a

weaponized version, and, therefore, they were able to ac-cept that public health professionals were learning as thecrisis evolved:

I’ve got to tell you in their defense, I’m not sure theycould have done anything differently because of theirown lack of knowledge. I would wager to say that theyprobably did the best they could do with what they had towork with.

In the midst of the crisis, public health agencies had toquickly field many staff. Participants perceived some ofthe public health professionals as being uninformedabout anthrax. When public health responders were un-aware of the published information postal workers hadfound during a search for information, participants de-scribed them as less than credible. For example, one par-ticipant offered: “I didn’t feel like they were fully trainedeither, because I was discussing some stuff I had read inarticles about anthrax, [and] they didn’t even know any-thing about it.” For many participants, the inability ofpublic health professionals to answer questions fueledsuspicion. Some participants assumed that federal agen-cies worked together and shared information, and manyparticipants specifically stated their assumption thatCDC had all the military’s information on anthrax.

As evolving events led to new scientific findings, pub-lic health professionals issued new recommendations.How professionals managed the uncertainty affected par-ticipants’ perception of credibility and competence. Oneparticipant lamented, “They would run out of the meetingand get on their cell phone and come back with differentinformation.” A less common recollection was: “It wasnew to them so as they got updated, they updated us, sothat part I felt was trustworthy.”

Critically important to participants was the way publichealth professionals handled the uncertainty. One workersaid, “Even if you don’t know information, you should letpeople know that you’ll get back to them when you haveinfo and follow through.” However, many participantslamented the lack of follow-through: “When they came tothese meetings with us, they could never really answer thequestions. They would always say well, we’ll get back toyou on that, and they would never get back to us.”

Openness and HonestyParticipants repeatedly stressed the need for honesty

and openness. This was a common refrain: “ . . . the firstthing you have to do is be honest. . . . [W]hen a crisiscomes, they think about being honest and tell the workerwho might be affected that it can cause them bodily harm,can cost our lives. They rather cover up than be honest.”

A common myth about communication is that full dis-closure of information in a crisis will lead to public panic.

QUINN ET AL.212

Participants rejected this assumption, suggesting that peo-ple react rationally when given proper information: “It’s alot easier to stop panic, stop me from panicking, when I’minformed than it is to stop me from panicking when I’mnot informed because then I would be going on rumors.. . . ” In fact, participants advocated that having more in-formation increased their ability to make decisions forthemselves: “They are underestimating citizens. If you behonest with people, people have a tendency to take risk aslong as they know what they are dealing with and whatcan happen to them if they deal with it.”

Caring and EmpathyMany participants did not perceive public health profes-

sionals as caring, concerned, or empathetic. One workersaid, “I felt that there was nothing or nobody that was con-cerned about me.” Others lamented that public health pro-fessionals were taking information from them and collect-ing “data” without inquiring about their concerns andwell-being: “So, if there was more communication with usfrom them, from all the different agencies, we would havefelt better. We would have felt more like we were beinglooked after—that someone gave a damn.” Many de-scribed their need for empathy and reported that this lackof concern contributed to diminished trust in professionals.

Many felt that the lack of a caring response was relatedto the USPS’s valuing economic interests over humanlife. One worker elaborated: “It seemed like they [USPS]were worried more about moving the mail, makingmoney, than finding out if the building was safe for theworkers.” Consequently, that may have increased partici-pants’ need for public health professionals to demon-strate commitment to them: “We wanted to know thatyour dedication is to the public.”

TrustNumerous factors affected the extent to which partici-

pants trusted public health professionals. Some postalworkers linked a lack of caring and subsequent lack oftrust to race and class. Numerous postal workers at theBrentwood facility identified race as a factor in the dif-ferences in treatment of postal workers and Congres-sional staff: “We felt . . . that the people at the Capitolwere important but the black people over there at Brent-wood, which is 99.9% black, didn’t matter.” Workers inNew York echoed the sentiment. However, many work-ers across all three sites felt a sense of second-class citi-zenship regardless of race: “ . . . we felt like second hand. . . you know, compared to the people at the Capitol.”

As new discoveries led to changing statements and rec-ommendations, participants described declining trust inpublic health professionals. Early reassurances also con-tributed to later distrust as the information was proven tobe incorrect. One worker characterized this sentiment:

The information they gave us did not back the informa-tion that was given to us after they closed the building.Because they are saying it’s safe and now they’ve closedthe building. If you closed it on Friday, was it safe onWednesday and Thursday, or did you just not react intime? Trust is going down in that way.

In contrast, some participants, particularly in NewYork, acknowledged that they did trust public health pro-fessionals, but they also confirmed the information aboutanthrax with other sources, whether from the Internet orthe media. Others felt they had to trust what public healthprofessionals were telling them, “because I had nothingelse to go on and I would rather be safe than sorry.” Theneed to build new relationships during the crisis was animpediment to trust. One person reflected, “Trust is builton past experiences, so it is hard to trust someone whenyou don’t know them.”

Some participants suggested that lack of communica-tion contributed to rumors and conspiracy theories. Somethought that the attack was a test of the government’s ca-pacity to respond to a new terrorist threat. Others believedthat the attack was an attempt by the USPS to downsizethe postal workforce and facilities. One worker describedit this way: “I mean they [USPS] want to close our placedown. They want to farm some of the mail out.”

For many, trust in public health authorities is eroded tothe point of disrepair. This was a common sentiment: “Ifwe had somebody just admit that there were mistakesmade, then we may be able to open up to trust. But untilsomebody takes accountability, we trust nobody.”

DISCUSSION

The challenges of risk communication during the an-thrax attack included a rapidly evolving, complex situa-tion that occurred across multiple sites. This study ex-pands on previous studies17,18,21 to provide an in-depthunderstanding and rich insights from participants in threepostal facilities who experienced the crisis and offersguidance for crisis and emergency risk communication.

Multiple factors diminished the credibility of publichealth professionals and diminished trust. The partner-ship between public health agencies and USPS manage-ment negatively influenced the ability of public healthprofessionals to be seen as trustworthy sources. Early at-tempts to reassure, offer risk comparisons, and minimizerisk backfired. The perceptions that public health profes-sionals were not knowledgeable and, even more damningin the eyes of postal workers, not empathetic and caringwere devastating to the success of the communication ef-forts. Additionally, this study confirmed findings fromStein and colleagues,17 who found that early messages

COMMUNICATION DURING THE ANTHRAX ATTACK 213

about use of gloves and masks were, in fact, harmful asthe workers understood the inadequacy of such protec-tive measures.

A major lesson is that early communication with af-fected audiences is critical. It is possible that, had publichealth professionals conducted a communication triageas suggested by Chess and colleagues, postal workerswould have been identified earlier as a potential audi-ence.21 In the future, public health professionals must beable to do a rapid assessment of the social and politicalenvironment in which any audience experiences a threat.

First steps in a rapid assessment would include a seriesof critical questions asked of the first contacts the profes-sional meets: Who are priority audiences? Do they haveany special needs, or language or literacy issues that mustbe considered? What channels are appropriate?

Public health professionals also must reach out to anyaudience quickly, starting with informal interviews toask: Are there any unique environmental factors that maycreate differential risks? Who are leaders and partnersthat the audiences trust? What key issues should publichealth professionals know? How can we best work withyou? Reaching out to partners such as unions or other or-ganizations is another critical early step. While the initialcommunication during a crisis may not wait, all furthercommunications can be improved by the assessment.Such an assessment can contribute to triaging accordingto differential risk at the worksite.

This study also confirms earlier findings that workersperceived that their race and occupational class explainedthe difference in their treatment compared to congres-sional staff.17,18 Developing trust is critical to the willing-ness of any audience to work with public health profes-sionals during a crisis. Therefore, it is essential thatprofessionals understand the social and historical contextin which many minority postal workers will receive com-munication. This requires that professionals increasetheir own confidence in directly addressing distrust andhistorical issues such as the Tuskegee syphilis study.

The challenges of any terrorist event will include sig-nificant uncertainty and changing information. Like theparticipants in the Blanchard et al. study,18 participants inthis study reported that public health professionals didnot consistently explain why information changed or re-turn to workers with new information, which harmed thecommunication efforts. In the future, participants recom-mended that public health professionals acknowledgewhen their information is incomplete, explain what theywill do to find the answers, and then, most important, re-turn to them to report the new information. This recom-mendation is consistent with Chess et al.’s21 recommen-dation that public health agencies make the explanationof contradictions and mistakes a communication priority.

Although public health agencies created materials

specifically for postal workers, the finding that partici-pants gathered most of their information from the massmedia and colleagues was consistent with other stud-ies.18,21 This suggests they often did not receive messagesthat addressed their specific concerns or in the time framein which they needed them. In the future, while publichealth professionals should specifically ask the media,including smaller local media, to publish key informationpertinent to postal workers (or other key audiences), it iscritical that public health agencies develop multiplemeans of communicating about the specific risks postalworkers face, including disseminating written materialsthrough unions and routinely conducting meetings withworkers on all three shifts to answer their questions dur-ing a crisis. Participants also reported that they would ac-cess a website and hotline that includes messages forpostal workers. They suggested disseminating informa-tion about the website and hotline through the media, na-tional and local unions, and USPS management.

There are other lessons from the anthrax attack. ForCDC and local and state public health agencies, it is nec-essary to continually educate the public about the role,responsibilities, and limitations of their agencies in thecontext of a terrorist attack. Many participants reportedthat they want to know how agencies have changed theircommunication practices to be more responsive.

Postal workers emphasized that having professionalsdemonstrate empathy, concern, and caring throughout thecrisis, including acknowledging fear, anger, and otheremotions of postal workers, is absolutely necessary. Theyoffered these specific suggestions: be available on thescene early; have a location apart from USPS manage-ment’s supervision; continue to be visible and present, in-cluding visits to the three shifts of the workday; holdsmall group discussions; listen to people’s concerns; andhave two-way conversations in which postal workers canask questions. One postal worker asked for this: “Talk tome and ask me how I am doing and not just in the contextof collecting information.” Another postal worker de-scribed the quality he believes is important in the interac-tion with professionals: “It is important to speak and thinkfrom ‘the heart.’ ” This focus on caring and empathy is anew finding that did not emerge in the other studies.

The closing of several postal facilities around Washing-ton, DC, in March 2005 because of an anthrax scare andthe discovery of ricin in postal facilities in late 2003 andearly 2004 indicate that postal workers will remain vulner-able to bioterrorist activity. In crisis and emergency riskcommunication, we know that information will always beincomplete and rapidly changing. However, a primary les-son is that if public health professionals build trusting rela-tionships with postal workers, union locals, and others, ifthey show concern and empathy by being available di-rectly to workers, and if they earn credibility by being

QUINN ET AL.214

knowledgeable, the likelihood that postal workers can ac-cept the incomplete and imperfect nature of the evolvinginformation is higher. It is essential for public health pro-fessionals to make amends now, to start to repair damagedtrust, and to develop mechanisms to reach postal workerswith timely, relevant, and credible crisis and emergencyrisk messages. Sorenson states it simply when he says that“ . . . it will take more than just good communication toshape public response to a bioterrorist event. It will also becrucial to have public trust and confidence.” 9(p231)

ACKNOWLEDGMENTS

The authors wish to thank the postal workers whokindly shared their time and insights with us. This studywas supported by grant #S2136-21/21 from the CDC/ASPH Cooperative Agreement. Dr. Quinn was also sup-ported, in part, by the EXPORT Health Project at theCenter for Minority Health, GSPH, University of Pitts-burgh, NIH/NCMHD Grant No. 5P60MD000207-03.This research was approved by the Institutional ReviewBoard (020948) at the University of Pittsburgh.

REFERENCES

1. Jernigan DB, Raghunathan PL, Bell BP, et al. Investigationof bioterrorism-related anthrax, United States, 2001: Epi-demiological findings. Emerg Infect Dis 2002;8(10):1019–1028.

2. Greene CM, Reefhuis J, Tan C, et al. Epidemiological in-vestigation of bioterrorism-related anthrax, New Jersey,2001. Emerg Infect Dis 2002;8(10):1048–1055.

3. Tan CG, Sandhu HS, Crawford DC, et al. Surveillance foranthrax cases associated with contaminated letters, NewJersey, Delaware, and Pennsylvania, 2001. Emerg InfectDis 2002;8(10):1073–1077.

4. U.S. Centers for Disease Control and Prevention. Evalua-tion of bacillus anthrax contamination inside the Brent-wood mail processing and distribution center—District ofColumbia. MMWR Morb Mortal Wkly Rep 2001;50(50):1129–1133.

5. U.S. Centers for Disease Control and Prevention. Update—adverse events associated with anthrax prophylaxis amongpostal employees—New Jersey, New York City, and theDistrict of Columbia metropolitan area, 2001. MMWRMorb Mortal Wkly Rep 2001;50(47):1051–1054

6. Reynolds B. Introduction to crisis and emergency riskcommunication. In: Crisis and Emergency Risk Communi-cation. Atlanta, Ga: Centers for Disease Control and Pre-vention; 2002:1–12.

7. Covello V, Peters R, Wojtecki J, Hyde R. Risk communi-cation, the West Nile Virus epidemic, and bio-terrorism:responding to the communication challenges posed by theintentional or unintentional release of a pathogen in an ur-ban setting. J Urban Health 2001;78(2):382–391.

8. Shore D. Communicating in times of uncertainty: the needfor trust. J Health Commun 2003;8(suppl):13–14.

9. Sorenson J. Risk communication and terrorism. BiosecurBioterror 2004;2(3):229–231.

10. The Working Group on “Governance Dilemmas” inBioterrorism Response. Leading during bioattacks and epi-demics with the public’s trust and help. Biosecur Bioterror2004;2(1):25–40.

11. Ratzan S. Communicating risk: from crisis to calm. JHealth Commun 2005;10:103–104.

12. Peters R, Covello V, McCallum D. The determinants oftrust and credibility in environmental risk communication.Risk Anal 1997;17(1):43–54.

13. Trettin L, Musham C. Is trust a realistic goal of environ-mental risk communication? Environ Behav 2000;32(3):410–427.

14. Kasperson R, Golding D, Tuler S. Social distrust as a factorin citing hazardous facilities and communicating risks. JSoc Issues 1992;48(4):161–187.

15. Frewer L. The public and effective risk communication.Toxicol Lett 2004;149:391–397.

16. Trumbo C, McComas K. The function of credibility in in-formation processing for risk perception. Risk Anal 2002;23(2):343–353.

17. Stein B, Tanielian T, Ryan G, Rhodes H, Young S, Blan-chard J. A bitter pill to swallow: nonadherence with pro-phylactic antibiotics during the anthrax attacks and the roleof private physicians. Biosecur Bioterror 2004;2(3):175–185.

18. Blanchard J, Haywood Y, Stein B, Tanielian T, Stoto M,Lurie N. In their own words: lessons learned from those ex-posed to anthrax. Am J Public Health 2004;95(3);489–495.

19. U.S. General Accounting Office. U.S. Postal Service: IssuesAssociated with Anthrax Testing at the Wallingford Facility.Washington, DC: GAO; 2003. Available at: www.gao.gov/new.items/d03787t.pdf Accessed June 12, 2003.

20. Mullin S. The anthrax attacks in New York City: the “Giu-liani press conference model” and other communicationstrategies that helped. J Health Commun 2003;8(suppl):15–16.

21. Chess C, Calia J, O’Neill K. Communication triage: an an-thrax case study. Biosecur Bioterror 2004;2(2):106–111.

22. Morse J. Qualitative generalizability. Qual Health Res1999;9(1):5–6.

Address reprint requests to:Sandra Crouse Quinn, PhD

Associate ProfessorDepartment of Behavioral and

Community Health SciencesGraduate School of Public Health

University of Pittsburgh230 Parran Hall

130 DeSoto StreetPittsburgh, PA 15261

E-mail: SQuinn@pitt.edu

COMMUNICATION DURING THE ANTHRAX ATTACK 215