Planning for Mass Prophylaxis with the Weill/Cornell Bioterrorism and Epidemic Outbreak Response Model (BERM)
Hospital and health system planners face new threats to the health and welfare of the public that few could have anticipated only three years ago. The anthrax attacks of 2001 in the United States, the global SARS epidemic that started in China in 2002, and the monkeypox outbreak of 2003 in the United States all highlight the need for community-wide response strategies to bioterrorist attacks or natural epidemics. Unfortunately, few tools are available to help public health and emergency management planners understand the logistical and staffing needs of a large-scale prophylaxis campaign designed to cover a specific target population in a limited time frame. Researchers at Weill Medical College of Cornell University in New York City used operations research principles to develop the model you are about to use. This model is part of a growing “toolbox” for planning mass prophylaxis strategies for bioterrorism and epidemic outbreak response.
The purpose of this interactive model is to allow you to “think with numbers” as you go about formulating realistic response plans for their local jurisdictions. Modeling forces critical examination of assumptions about prophylaxis strategies and about the availability of resources such as staff and potential prophylaxis clinic sites. Estimates derived from this model should be viewed as one type of data among many that may be useful in formulating response plans (other data might include, for example, previous local experience with immunization campaigns or results of training exercises for bioterrorism response).
As with any model, the accuracy of outputs provided here depends on the quality of the underlying data on which they are based. For example, the patient processing time estimates used in this model have a large impact on outcomes (to demonstrate this, observe the change in overall staffing estimates for a given scenario under slow, baseline, and fast processing times). In order to minimize spurious variability in outputs, the model provides three pre-set choices for processing times and three pre-set choices for disease prevalence. On the other hand, in order to allow maximal individualization of outputs for a particular community, the model also includes a section that allows advanced users to customize every element of these baseline scenarios.
Nathaniel Hupert, MD MPH Department of Public Health Weill Cornell Medical College