Adaptive Response Metric (ARM)
The “Adaptive Response Metric” (ARM) measures changes in resource consumption over the course of a response as a Local Health Department (LHD) adapts from normal, to disaster, and eventually back to normal levels. The ARM measures adaptation at the level of each agency function and aggregates the scores to achieve an LHD profile, recognizing that agency functions may not be uniformly engaged. The ARM generalizes in two basic ways: First, it may be applied to LHDs in different locations or, it may be applied to the assorted public health agencies in a give locale. Second, with continued application, analysis, and testing, the ARM will measure responses to emergencies or disasters of any type or scale. The ARM enhances after-action reporting by providing a standard measurement of resource consumption across agency functions (i.e., divisions or departments), distinguishing those actually involved in an emergency response from those maintaining routine operations. The ARM offers the potential for supplementing after-action reporting methods—which may be lengthy, inconsistent, and requiring subjective interpretation—with standard, validated, quantitative methods that apply across agencies, disaster types, and levels of severity. This generalizability greatly increases the availability of reliable and valid information for decision-makers and policy makers considering improvements in preparedness and emergency response.
Local Health Department Characteristics Database (LHD Database) and the Geospatial Area and Information Analyzer (GAIA)
The LHD Database is based on a developed taxonomy that classifies almost 3,000 local health departments by key attributes of governance (legal and policy authority) and administration (local decision-making). These attributes are central to the ability of local public health to prepare for, respond to, and recover from disasters and emergencies.
Through the GAIA pilot study, an interactive, Web-based application of the LHD Database is being created and becomes a way for public health officials, researchers, and the general public to explore this important dataset visually and geographically. Jurisdictions can be analyzed both by preparedness indicators (funding, FTEs, presence/absence of a preparedness coordinator, proximity to high-risk targets, etc.) and more general health indicators (population, general and specific causes of death, level of poverty, etc.). This analysis improves understanding of the similarities and variations of public health structure at the local level, allowing policy makers and planners to better understand capacity and capability and aiding in decision-making about allocation of resources and ways to enhance overall performance.
Legal Networks Analyzer (LENA)
To improve nuclear emergency preparedness and response capacity in less experienced yet nonetheless vulnerable states, the Legal Network Analyzer (LENA), a web-based applet, was created. LENA allows policymakers to visualize the laws and regulations directed in the National Response Framework (NRF) or present in more experienced states yet missing in their state laws. The text of FEMA's NRF nuclear incident annex was coded to identify the specific actions and directives our nation's chief Emergency Management Agency has given to prepared for and respond to nuclear incidents in the United States. The NRF nuclear incident directives were then compared to the nuclear incident directives given by each state legislature. The laws were mapped and legal networks created to visually demonstrate marked state variability in preparedness and response capacity for nuclear/radiologic incidents.
The draft interactive applet can be accessed at: www.phdl.pitt.edu/LENA
Public Health System Emergency Preparedness and Response Law Database
Each agent of the public health system (PHS) is directed by laws that are unique to that organization's role within the system. To determine how each agent, with independent legal duties, is directed to function, qualitative research methods were used to code the text of the laws directing 42 PHS agents. The laws were coded to denote the actions, purposes, goals, and conditions each agent was directed to undertake by their respective legislature. Employing quantitative research methods and SPSS and UCINET software, the legal networks undergirding our nation's PHS emergency preparedness and response capacity were demonstrated and analyzed. The states selected for study (Alaska, California, Florida, Maryland, Pennsylvania, Texas, and Wisconsin) represent 5 of the 10 Health and Human Services regions of the country, vary in population density (urban, rural, and frontier states), and demonstrate differing levels of experience with a variety of disasters (snow storms, radiologic/nuclear incidents, floods, earthquakes, hurricanes, and fires).
Analysis of the networks created by the laws directing state emergency preparedness and response activities revealed extreme variability in the legal structure undergirding our nation's preparedness for and response to emergencies with public health consequences. Fewer than 400 statutes direct emergency preparedness and response in Alaska, with virtually no direction provided to more than a third of the PHS agents in the state.
Alternatively, close to 6000 statues direct the entire span of PHS agents in California. Further, network analysis tools revealed marked variation across states in terms of PHS legal network structure, centrality, density and in and out degrees. The project’s systematic coding of the laws and mapping of state PHS emergency preparedness and response capacity enables policy makers to visualize network strengths and weaknesses and enables them to draw upon the disaster experience and legislative response of other states, and thus improve emergency preparedness and response capacity in their state.
The coded sections of state statutes and regulations, as well as all coded sections of the National Response Framework Nuclear Incident Annex, were entered into a publicly accessible database that is searchable by state and/or PHS agent. The database may be accessed at: http://www.phasys.pitt.edu/default.aspx. A PDF instruction sheet on how to use the database can be accessed at: Law Database Instructions. A PDF of the codebook used to develop the database can be accessed at: PHASYS Arm 2 Code Book 11.18.12.
Inter-Region Epidemic Dynamics (IRED) Model
The IRED simulates the spread of contagious disease throughout a large multi-patch region like the United States. For any geographic distribution of observed cases, the IRED model quickly generates an ensemble of stochastic realizations and computes the mean and standard deviation of incidence for every patch. Spatially heterogeneous epidemic severity can then be assessed to provide relevant estimates for local contingency planning. The IRED model can estimate whether an epidemic beginning in one patch will spread to particular other patches in a statistically credible way. The IRED model is ideal for assessing the impact of government-imposed travel restrictions or endogenous changes in travel behavior. It permits short-term epidemic forecasts to be made in a mathematically rigorous way.
Interactive Large Scale Agent Model (ILSAM)
The ILSAM is a powerful interactive desktop simulator. The original Large Scale Agent Model is a flexible model of disease transmission with the ability to simulate the entire population of the United States in around 10 minutes. The ILSAM is rule-based and avoids highly mathematical representations, making it extremely user-friendly. This allows individuals unfamiliar with any type of modeling to easily interact with an agent model. An agent based model allows a completely synthetic population to be created and to act as real people do. Epidemic dynamics are then studied by watching how disease spreads though the computerized population.
The ILSAM is a redesigned version of the Global Scale Agent Model (a distributed high performance agent model) that retains many of the capabilities of its high performance predecessor, but requires neither highly specialized computing platforms (clusters) nor programming skill. Removing these requirements has produced a new tool that enables anyone to execute richly specified epidemic simulations. It is hoped that the ILSAM will not only be a valuable tool for evaluating disease interventions but also a valuable tool for government training at various levels (national, regional). The models are transparent and "rule based" (e.g., each day, all children go to school, where they contact K other randomly selected children; at night, all families are at home together, resulting in M contacts). All action is depicted graphically, as if users were looking down on the social space from above, watching agents move to and from the various social units (e.g. homes, schools, workplaces, clinics), changing colors as they progress through the phases of the disease.
The ILSAM will be uniquely helpful when evaluating pharmaceutical and non-pharmaceutical policy because candidate decisions can frequently be implemented directly in an agent model. For example, the impact of school closures can be easily studied because ILSAM includes all schools in the United States and how many students attend each of these schools. The impact of school closures has been shown to depend on variables like: duration of closure, the level of sickness that prompts a closure, and whether school closures are synchronized locally. Effective vaccination strategies can also be studied using agent models. Optimal vaccination strategies will depend on many real-world logistical constraints. Agent models are well suited to including these real-world constraints because agent models are capable of extremely high resolution.
Interactive Models for Public Health
The Interactive Models for Public Health are local models (written in NetLogo) that provide a more granular, local look at the effect of various preparedness activities. These models have a user-friendly interface permitting immediate experimentation and visual feedback in real-time running on a laptop. Any user can click on a node to give it information, and then watch the information flow through the networks. Similarly, any user can give any node resources and watch it flow through the network. Depending on which model, there are numerous variables that can be manipulated.
The various interactive draft models for public health include:
Decisions Support Dashboard
The Decisions Support Dashboard is a working prototype that estimates the likely impact upon a hospital’s capacity to deliver medical services such as: surgical operating rooms, intensive care services, emergency room functions, and nursing station assignments, if any of the key engineering services is disrupted. The current status of key technical functions are displayed: water, steam, chilled water, electrical power, gas/oil, communications, medical gases/vacuum system, pneumatic tube system and their change in status in real time to inform hospital administrators of the estimated time available to adapt medical services to altered capacity of engineering services. The Dashboard is designed to link to external utilities such as water and electrical power to show how a disruption caused by a hazardous event in community will affect operation of hospitals.