Following the September 11 terrorist attacks and the ensuing anthrax outbreak, the majority of the emergency physicians we surveyed in Pennsylvania had received patient requests for anthrax prophylaxis; a quarter of these physicians prescribed antibiotics for these patients. Physicians that reported patients' requests for antimicrobial prophylaxis were distributed across the state, suggesting that patients' search for protection against anthrax was widespread. Our results are consistent with data on specimens submitted to Pennsylvania public health laboratory officials for B. anthracis
analysis. During October-December 2001, the Pennsylvania Bureau of Laboratories received approximately 1400 specimens including white powder, environmental swabs, and letters from counties in eastern and western Pennsylvania districts. Of these, 27 (or about 18 requests per 100,000 population) came from Cambria County in the western district (PA DOH: unpublished data). There were 11,063 anthrax-related telephone inquiries received from October 8 to November 11, 2001 by the Centers for Disease Control and Prevention's (CDC) emergency operations centers; queries originated from all states and one US territory. Most of these calls were from members of the public concerning anthrax vaccines (≈58%), suggesting that search for protective measures against anthrax was widespread across the United States [11
Other studies have documented an increased use of antimicrobial agents that was temporally related to the anthrax outbreak; this use could not be ascribed to that recommended by the CDC. For example, a recent national study reported that approximately 160,000 more ciprofloxacin and 96,000 more doxycycline prescriptions were written in 2001 compared to 2000 [12
]. When other investigators compared ciprofloxacin utilization in 2001 with 2000, they found that it declined for all months except October 2001 when ciprofloxacin utilization increased 9.8%. They also found that the increase was not limited to areas where anthrax cases had occurred, suggesting that many Americans sought antibiotic prophylaxis [13
Of the physicians reporting that they had prescribed antibiotics for anthrax prophylaxis, the majority used ciprofloxacin. This finding is consistent with other studies [12
] and is likely because the initial CDC guidelines recommended ciprofloxacin prophylaxis for B. anthracis
exposure until susceptibility results were known [4
]. When tests showed that B. anthracis
isolates recovered from patients involved in the anthrax attacks were susceptible to other antibiotics, public health officials indicated that doxycycline might be preferable over ciprofloxacin [2
]. While both drugs are approved for postexposure prophylaxis [14
], the rationale for favoring doxycycline was to prevent ciprofloxacin resistance in more common bacteria.
Unfortunately use of antibiotics has inherent risks and costs and optimizing benefits is especially difficult in the midst of bioterrorist events. Consequences of antibiotic treatment of unexposed individuals include adverse drug reactions, increased risk of antimicrobial resistance, depletion of antibiotics, and monetary costs [15
]. Furthermore, use of emergency departments for sporadic distribution of prophylactic antibiotics to persons presenting with self-identified risk appears inefficient. It is unclear whether these persons can be adequately managed in emergency departments without the support of public health and law enforcement officials.
Public health response likely influences demand for and outcomes associated with antibiotics requests during bioterrorism attacks. When we asked physicians to suggest on what health departments could do to reduce the influx of patients to the emergency departments, they cited official communications to make the public "less worried." In Illinois, a surge in environmental samples received by public health officials for anthrax tests was associated with both media reports of anthrax cases in other states and a specific announcement on October 29, 2001 by the US attorney general and the FBI director. The announcement asked US citizens and law enforcement agencies to be on the "highest alert" based on "credible information" [16
]. Lessons learned from the 2001 anthrax attacks in New Jersey suggest that communities in which the public health sector and clinicians have a strong working relationship are better prepared to meet mass prophylaxis needs [17
]. Similarly, lessons learned in New York City during the same outbreak demonstrated the benefits of advance logistical planning for mass postexposure prophylaxis including an antibiotics distribution site and clear eligibility criteria [18
We acknowledge some limitations to our results. First, as in any survey, these data are subject to non-response bias. But the 47% response rate is comparable with other telephone surveys conducted among physicians in general [19
]. In addition, responders and non-responders had similar baseline characteristics, suggesting that these groups were comparable [9
]. Second, the study is limited to types of antibiotics prescribed and cannot be used to estimate dosage, number of pills allotted to these patients, costs, or compliance to treatment for perceived or real anthrax exposure. Third, the indications for prophylaxis were not studied. While it is not certain, it is likely that at least the vast majority of the antibiotic perscriptions found in this study were outside indications described in public health guidelines. It is plausible that some patients sought prescriptions for storage; it is also likely that they at least initiated the antibiotic course.