Setting and Background

On October 15, 2001, in an office of the Washington, D.C., U. S. Capitol complex, an envelope addressed to Senator Tom Daschle, intentionally contaminated with B. anthracis spores, was opened. This event occurred 2 weeks after a report from Florida of the first-ever inhalational anthrax cases related to envelopes containing B. anthracis spores; those cases occurred in employees of a media company (3). The Washington, D.C., Department of Health (DCDOH), Office of the Attending Physician, U.S. Capitol, and the Centers for Disease Control and Prevention (CDC) immediately initiated a multiagency public health response and epidemiologic investigation (7). Enhanced surveillance activities for inhalational anthrax in the national Capitol area were established through a cooperative effort of the DCDOH, Virginia Department of Health, Maryland Department of Health and Mental Hygiene, and CDC.

The epidemiologic investigation determined that the B. anthracis–contaminated envelope addressed to Sen. Daschle was processed on October 12 at the DCPDC before entering the Capitol mail distribution system. Late on October 19, a DCPDC employee was admitted to a Virginia hospital with a diagnosis of suspected inhalational anthrax. The CDC team visited the DCPDC on October 20. The suspected case-patient worked in an area of the DCPDC where the envelope had not been processed; he also worked in a second mail facility in Maryland. The diagnosis of inhalational anthrax was confirmed on October 21 (1,2). The DCPDC and the second mail facility in Maryland were closed on October 21. On October 20–22, three additional cases of suspected inhalational anthrax were identified in the DCPDC employees; two of these patients died (Table 1). B. anthracis grew from blood cultures from all patients within 24 hours. B. anthracis was confirmed by B. anthracis–specific polymerase chain reaction assay at CDC (2).

A second envelope with B. anthracis spores, addressed to Sen. Patrick Leahy, was identified on November 16. This envelope was recovered from a sealed drum containing U.S. Capitol mail quarantined on October 17, 2001.

Postexposure Prophylaxis

After confirmation of the first case of inhalational anthrax in a DCPDC employee, antimicrobial postexposure prophylaxis (PEP) was recommended to all DCPDC employees and visitors to the nonpublic mail-processing area (3,8). DCPDC employees who had been absent from work >24 hours in the past 7 days were contacted to identify any additional cases and inform workers of the recommendation for PEP. Beginning October 21, workers were given a 10-day supply of antimicrobial therapy, pending further investigation. DCPDC employees returned to the public health department antimicrobial agent distribution centers to receive an additional 50-day supply of antimicrobial therapy. All DCPDC employees were offered free medication from the U.S. National Pharmaceutical Stockpile at D.C. General Hospital through a cooperative effort of the DCDOH and the U.S. Public Health Service. Employees could choose to obtain appropriate medication from other sources. The United States Postal Service (USPS) notified employees from the DCPDC about the recommendation for postexposure prophylaxis and urged them to comply. Information on the symptoms of inhalational anthrax, the biology of B. anthracis, and possible adverse effects from antimicrobial agents was distributed to postal workers. The number of employees who obtained antimicrobial therapy from D.C. General Hospital, the Virginia Department of Health, and Maryland Department of Health and Mental Hygiene was recorded.

Postal System Assessment

In collaboration with the USPS and the Postal Inspection Service, we assessed routine mail-handling procedures and reviewed the path of the two envelopes that were known to contain B. anthracis spores. From unique envelope markings, postal inspectors determined the time of automated envelope processing and the machinery used during processing. To establish the number of employees potentially exposed during the passage and processing of the two envelopes, DCPDC employee work zone locations, job descriptions, and assigned work shifts were obtained from USPS administrative data.

Case Exposure Histories

We interviewed surviving case-patients and close associates of those who died by using a standard exposure questionnaire. Case-patients were assessed for job description, work and break locations, travel and medical history, and potential exposure to natural reservoirs of B. anthracis spores. Timecard logs established exact times of work during October 11–21.

Environmental Assessment

Beginning October 23, the DCPDC facility was sampled extensively with a combination of surface wipes, surface vacuum samples, and air vacuum samples, reported in detail elsewhere (9–11).

Nasal Swab Cultures

Nasal swab cultures from the DCPDC employees and those who reported visiting that facility during the period October 10–21 were obtained on October 21–22 during the distribution of antimicrobial therapy. Specimens were processed by standard microbiologic methods at the Maryland Department of Health laboratory (12).

Serologic and Exposure Survey

We conducted a survey to evaluate occupational exposures of workers and determine whether there was evidence of immunologic response to B. anthracis protective antigen. Exposure histories and serum samples were obtained from a convenience sample of DCPDC employees who went to D.C General Hospital on October 29–30 for their additional 50-day supply of antimicrobial therapy. Each participant was asked to allow a serum sample to be collected and to be individually interviewed with the standardized exposure questionnaire used for case-patients. Informed consent was obtained from all participants.

One blood sample was obtained from each participant. The serum was separated and stored at 4°C. Anti-protective antigen immunoglobulin G (anti-PA IgG) antibody concentrations in serum specimens were determined by a quantitative enzyme-linked immunoassay described in detail elsewhere (13).

Comparison of Case-Patients and Survey Participants

We compared exposure histories and underlying diseases of the case-patients with the sample of surveyed workers to clarify factors that may have contributed to the four cases of inhalational anthrax at DCPDC. Data from the standardized exposure questionnaire from the DCPDC cases and the other sampled employees were compared by a case-control analysis with two-tailed Fisher exact tests for dichotomous variables or the Wilcoxon signed-ranks test for continuous variables; p values of <0.05 were considered significant.

Postexposure Prophylaxis

Of 2,403 employees at the DCPDC, 1,870 (78%) were recorded as receiving a 50-day supply of antimicrobial therapy at DCDOH, Virginia Department of Health, or Maryland Department of Health and Mental Hygiene postexposure prophylaxis distribution centers. Five members of the CDC team received PEP.

Postal System Assessment

Figure 1

Figure 1. Floor map of the Washington, D.C., Postal Processing and Distribution Center with the known locations of the two B. anthracis–contaminated envelopes (gray diamonds with letters) and work locations of case-patients (black...

The DCPDC is a 500,000-square foot facility (Figure 1). Approximately 59 million pieces of incoming mail were processed at the DCPDC during October 11–21. The two contaminated envelopes entered the DCPDC on the evening of October 11 or early morning of October 12 in a tray of envelopes originating at the processing and distribution center in Trenton, New Jersey. This tray was taken from the dock (Figure 1, point A) to a large tray-sorting machine (Figure 1, point B) and then moved to a high-speed envelope-sorting machine known as a delivery bar-code sorter (DBCS). The DBCS moves up to 30,000 letters per hour into a series of bins for subsequent distribution. At the end of each work shift, the DBCS is cleaned by a procedure that blows compressed air (70 lbs per square in) into the machine. Unique processing markings on the two envelopes showed that DBCS number 17 (Figure 1, point C) sorted both envelopes on October 12 between 7:05 and 7:30 a.m. The two letters appeared to be processed within minutes of each other.

The envelope addressed to Sen. Daschle was sorted into a bin destined for the U.S. Capitol, taken out of the DBCS number 17, and moved to the government mail section (Figure 1, point D). The government mail section handles distribution of all letters to U.S. government addresses in the metropolitan D.C. area. Routinely, in a process known as riffling, envelopes are flipped through individually for manual confirmation of appropriate sorting. The envelope addressed to Sen. Daschle transited the government mail section on October 12 between 7:30 a.m. and noon, at which time it was dispatched from the loading dock (Figure 1, point E) to the U.S. Capitol’s mail distribution facility.

The envelope addressed to Sen. Leahy was incorrectly sorted as destined for the U.S. State Department, State Annex 32, which has an independent small mail-processing facility in Virginia. The exact path of this envelope is unclear from October 12 to 17. On October 17, the Federal Bureau of Investigation quarantined all remaining U.S. Capitol mail and placed it into sealed drums for further investigation; the Leahy envelope was found in one of these drums on November 16. The letter appeared to be leaking. Routine procedures for redirecting incorrectly sorted envelopes destined for a U.S. government address usually involve employees in the DCPDC government mail section. If routine procedures had been followed, and the envelope were recognized as incorrectly sorted, the envelope would have been manually redirected in the DCPDC government mail section in the period October 12–16.

A fifth case of suspected inhalational anthrax in a postal worker in Virginia was reported on October 25, 2001. The case-patient worked in State Annex 32 (3). Whether the envelope to Sen. Leahy remained in the DCPDC or transited through the State Annex 32 is not known. Mail destined for the State Annex 32 was sorted at the same time in DBCS number 17 as the envelope addressed to Sen. Leahy. Environmental sampling results from State Annex 32 showed widespread contamination with B. anthracis spores, similar to the DCPDC.

Of the 1,961 employees of the DCPDC nonpublic mailroom area, 610 (31%) were assigned to work in one of the same work zones as the four case-patients from this facility. During the time of the two envelopes’ passage and processing through the DCPDC, approximately 108 (6%) worked in the main processing area with the DBCS machines between 12 a.m.–12 p.m. on October 12, and 87 (4%) employees worked in the government mail section (Figure 1). Two of the four case-patients worked in one of these work zones. The attack rate for inhalational anthrax in these combined areas was 1% (2/195).

Exposure Histories

The clinical characteristics of the case-patients have been described.(2) All four case-patients from the DCPDC were African-American; case-patient 5 from the mail-processing facility for the State Department was white (Table 1). Only two case-patients had underlying medical conditions. Case-patient 3 had adult-onset diabetes mellitus and a 30-year history of sarcoidosis, although the patient was not on medication for either condition (14). Case-patient 4 had a diagnosis of asthma and was periodically treated with bronchodilators.

Figure 2

Figure 2. Comparing the time period that the case-patients were at the Washington, D.C., Postal Processing and Distribution Center (solid black bars) to the time period that the two envelopes containing B. anthracis...

Two of the four case-patients from DCPDC worked within several meters of the path of the processed envelopes (Figures 1 and 2). Only one case-patient routinely worked directly with high-speed envelope-sorting machinery, including routine overtime on DBCS number 17. At the time the two contaminated envelopes were sorted in the DCPDC by DBCS number 17, only case-patients 2 and 4 were physically in the DCPDC facility (Figure 2). However, case-patient 1 returned during the window of time when DBCS number 17 was cleaned by blowing compressed air into the machine, between 8:00 a.m. and 9:40 a.m. Case-patient 3 returned to work in the government mail section (Figure 1, point D) at 8:00 p.m. on October 12.

Environmental Sampling

Diffuse environmental contamination with B. anthracis was found throughout the nonpublic mail-processing area of the DCPDC, particularly on DBCS number 17 and in the government mail section of the facility (9–11). In addition, two supply air ventilation diffusers, located above the area where two of the case-patients worked, were contaminated with B. anthracis spores (10,11). None of the samples taken from the public area of the facility were positive for B. anthracis spores.

Nasal Swab Cultures

Nasal swab cultures from 3,110 DCPDC employees and visitors, collected 9–10 days after the two envelopes were processed at the DCPDC, were negative for growth of B. anthracis.

Seroprevalence and Exposure Survey

On October 29–30, a total of 1,657 employees and visitors to the DCPDC went to D.C. General Hospital to receive additional antibiotic supplies. Of these, 784 (47%) were asked to participate in the survey; 224 (29%) of 784 DCPDC employees participated. Serum samples were obtained from 202 (94%). None of the 202 serum samples had significant detectable specific IgG antibody concentrations of anti-PA IgG, including the three participants who reported a remote history of anthrax vaccination.

The routine work activities of case-patients were also relatively common for the surveyed DCPDC workers (Table 2). Fifty-four percent reported that they manually sorted mail, and 39% reported that they riffled mail. Seven percent of survey participants reported that they riffled mail on October 12. Few employees reported the use of masks (7%), although 47% of survey participants and 50% of case-patients reported using gloves.

Comparison of Case-Patients and Survey Participants

Differences in underlying medical conditions or workplace exposures between the DCPDC case-patients and the survey participants were not statistically significant (Table 2). With sarcoidosis included as a chronic lung disease, more case-patients had chronic lung disease than did survey participants (50% vs. 9%; odds ratio 9.65; 95% confidence interval 1.29 to 72.2; p=0.01). None of the case-patients currently smoked cigarettes, compared with 24% of the participants. Specific mail-handling activities such as manually sorting mail or working on a sorting machine also did not differ. No case-patients and few (10%) of the serosurvey participants handled bulk mail.