Data for this paper come from four related projects conducted from 2003–2006. Taken together, these projects involved developing, conducting, and evaluating 31 tabletop exercises with state and local health departments of different sizes and structures in 13 different states across the northeast, south, mid-west, and west regions of the country (Table ). Participating health departments did not incur any expenses through their involvement in these exercises other than the staff time required to participate.
Descriptive Characteristics of Participating Health Departments
Two of these projects, one in California and the other in Georgia, involved the conduct of exercises in multiple jurisdictions in the same state. In California, the Little Hoover Commission, a bipartisan, independent state body, asked RAND to assess California's public health infrastructure. A key component of the project, described in greater detail elsewhere [16
], was the development of a tabletop exercise that simulated a smallpox outbreak. This exercise was conducted in seven local health departments across California. In Georgia, RAND collaborated with the Georgia Division of Public Health and the Rollins School of Public Health at Emory University to develop, conduct, and evaluate a series of tabletop exercises focusing on different biologic agents in seven local health departments across Georgia, as well as one exercise focused at the state level.
The two remaining projects were funded by US Department of Health and Human Services (HHS) and involved the participation of multiple local health departments. The first project involved developing ten different tabletop exercise templates and formats focusing on the local public health response to bioterrorist agents. These were tested in 13 local health departments in 12 different states. The second project involved developing a tabletop exercise to examine the interface between local health departments and health care systems in a hypothetical influenza pandemic. This exercise was tested in three local health departments in different states. Greater detail on the structure for these tabletop exercises as well as the tabletop exercise templates themselves can be found elsewhere [7
All exercises focused on at least one of three related objectives: training, relationship-building, and evaluation. The structure and design of the tabletop exercises varied from project to project because their objectives were somewhat different. The key domains covered are outlined in Table . The level of facilitator involvement varied with the exercise objectives. At one extreme, the facilitator's role was limited to introducing the exercise scenario and periodically interjecting updates. During these exercises, the participants were encouraged to lead the discussion themselves, based on their respective roles in their agency or organization. At the other extreme, the facilitator took a very active role by leading the discussion and interjecting questions or prompts. In between were exercises in which the facilitator turned the discussion over to participants but occasionally joined the discussion to request clarifications from the participants or assure that issues critical to the exercise objectives were discussed.
Tabletop Exercise Design Variability
Despite these differences, all of the exercises shared common elements, including: evolving hypothetical scenarios, facilitated group discussions, and some level of collective decision making by participants emphasizing the role of local health departments in recognizing and initiating a response to an emergency. The scenarios typically began with a single case report or series of case reports that heralded a nascent disease outbreak and required a public health assessment. These situations exercised the internal communication and coordination across disciplines within health departments as well as the communication and coordination with partner agencies and organizations such as health care facilities and emergency medical service agencies. Several exercises extended beyond this initial response and included scenarios that progressed days or weeks into an outbreak, requiring greater interactions between local- and state-level authorities and attention to health care surge capacity.
Every exercise concluded with a "hot wash" in which participants discussed their collective performance, identified strengths and weaknesses, and when relevant, related their performance to experience with actual outbreaks or crises. In the latter exercises, participants were prompted to develop an initial 'action plan' that addressed key vulnerabilities identified in the exercise. The facilitators subsequently generated a written "After Action Report" (AAR) that summarized the exercise experience and highlighted the observed strengths and areas for improvement. In addition, participants completed exercise evaluation forms. These consisted of a series of structured and semi-structured questions that asked participants to discuss what they learned during the exercise and to evaluate aspects of the exercise structure and conduct. For example, participants were asked to identify key gaps in preparedness that occurred during the exercise and to identify the most useful thing they learned during the tabletop exercise. The observations reported here are based on reviewing the after action reports, participant evaluations, as well as internal team discussions and consensus following the exercise debriefings.