| To Register SMDM Homepage | |
|
|
|
||||
Purpose: The incubation period and consequent epidemic curve caused by a large-scale anthrax exposure in a civilian population setting remains poorly understood and hotly debated. We investigated the impact of different epidemic curves to determine the minimum safe detection time and response time to achieve ≥97.5% disease prevention through mass antibiotic prophylaxis. Methods: Using a state transition model of medical prophylaxis for bioterrorism agents (MDM'03) we estimated the proportion of patients exposed to an infectious dose of aerosolized anthrax whose illness is prevented by timely antibiotic prophylaxis. Model assumptions include random mixing, medication efficacy (i.e., not antibiotic resistant) and patient compliance. We used two basic incidence curves for inhalational anthrax: a lognormal distribution of hospitalized cases derived from the 1979 Sverdlovsk outbreak (mean 2.398 (SD 0.713) with peak cases on day 7, from Brookmeyer, et al., Biostatistics, 2001) and a normal distribution estimated from combining the first two weeks of Sverdlovsk with the 11 inhalational anthrax cases in the 2001 U.S. outbreak (mean and peak day 8.3 (SD 3.0) using @Risk for Excel). We ran multiple sensitivity analyses varying the day of peak cases +/- 3 days for each curve. Results: Prevention of symptomatic inhalational anthrax in ≥97.5% of exposed persons is attainable only with extremely rapid agent detection and prophylaxis response (≤3d using lognormal incidence, ≤2d using normal incidence) and rapid completion of prophylaxis campaigns (≤5d for both distributions). Within this range of optimal response parameters, we found little difference in projected outcomes between the baseline lognormal and normal incidence distributions (Graph). However, substantial (>5%) divergence of outcomes occurs if anthrax incidence is more rapid than the baseline distributions predict (peak cases ≤day 6), with the lognormal distribution leading to consistently better prophylaxis outcomes. Conclusions: Prophylaxis campaign parameters and projected patient outcomes were similar using a previously published lognormal distribution and a newly-calculated normal distribution estimating the incidence of inhalational anthrax after an outdoor exposure. These results provide an evidence base for bioterrorism preparedness planning.
See more of Oral Concurrent Session B - Public Health 2
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)