Public Health Preparedness Modeling

Version 2.1 of eMedCheck is now available. eMedCheck is an electronic medication screening form that can be run on a PDA. Using this software, POD staff record basic information about each family member. The software uses decision rules to determine which medication each person should receive. It also records the results for later analysis.

Varun Santosh, a graduate student here at the University of Maryland, has done most of the work developing this application.

Version 2.1 features updates to the user interface, including this screen for entering the number of people in the family:

The same group at the University of Tübingen has a new version of InfluSim, which can “predict the course of an influenza epidemic in a fully susceptible population.”

The new version is a 20MB zip file that I downloaded and then extracted to a folder on my computer’s desktop. I then ran the executable and started using the program. The model allows the user to change a wide variety of parameters about the population, the disease, and treatment policies. The outputs include the number of people infected, the number of days of work lost, the number of hospital beds required, and the costs.

The extensive help includes a warning that the program “should not be used to predict the exact course of the next influenza pandemic.” Instead, its goal is help users understand how assumptions about the virus and the interventions used affect the epidemic. Waiting a few seconds after changing a parameter yields new results (shown as numbers and as graphs). The quick (though not instantaneous) display of new results should be useful to those who want to gain that insight.

Jason Matheny at the Johns Hopkins University Applied Physics Laboratory sent me a list of more computer models for public health emergency preparedness planning. I will posting my comments on these in future posts here.

First up is SimPox from a research team at the Department of Medical Biometry, University of Tübingen. SimPox runs as a Java applet. It simulates the spread of smallpox in a population. The user can change parameters about the disease (including its infectivity and latent period), the population (the initial fraction immune), and the intervention measures (such as surveillance and vaccination) and see charts with the predicted number of people affected each day of the smallpox attack. This includes the number of roaming cases, the number of cases detected, and the number isolated. If multiple simulations runs are used, then the results for each day can be are shown as a standard error around the mean, the minimum and maximum (over the simulation runs), and the individual simulation results (trajectories). The program also provides a text report with all of the parameter settings and numerical results, which one could copy and paste to save or print.

The Western New York Alliance Inc. Advanced Practice Center has created an Urban to Rural Evacuation Tool that predicts the movement of people from large cities to the surrounding counties.

The map-based tool presents a large set of relevant statistical data in a useful manner. It runs as a Java program in a web browser (I tested it using Internet Explorer). The first step is to choose a large city from a map of the lower 48 states. The choices range from New York and Los Angeles to Pierre, South Dakota, and Santa Fe, New Mexico. After a city is selected, the model shows a map of the counties surrounding that city. Moving the mouse over the counties shows key information about that county, including its population and the number of resources available to serve evacuees (grocery stores, hotels, restaurants, and so forth). One can also choose to select different views, so that each county is colored by the level of resources or the number of evacuees expected to move there under different scenarios.

The SNS Modeling Forum now has a blog called Looking Forward, which “is an attempt to broadly scan news and information sources for new technologies, policies and regulations that may potentially impact emergency response and public health in the future, and to provide links and commentary on those items.” Both resources are sponsored by the Strategic National Stockpile of the Centers for Disease Control and Prevention.

A primary use of models is to evaluate proposed or existing plans and designs. For example, how many people per hour can we vaccinate using this POD design? However, not all evaluation tools are mathematical models.

CDC has developed an Environmental Health Shelter Assessment Tool that can be used to evaluate a shelter (the kind used to house and feed people during or immediately after a disaster or an emergency. The tool is a checklist. Each item is some feature or attribute of the shelter that should be in place. For example, is there hot water? The items are grouped into areas such as facility, food, water, sleeping areas, sanitation, waste collection, medical care, child care, and companion animals.

Peyman Karimian, my research assistant, recently examined Master the Disaster, which was created by the Center for Public Health Preparedness, an Advanced Practice Center at the DeKalb County (Georgia) Board of Health. Peyman provided the following review:

The software is disaster designer software which creates real scenarios of disasters such as Anthrax, Botulism, Hurricane, Pandemic Flu, Plague, RDD (Radiological Dispersal Device) and Smallpox. The goal of the software is to provide real scenarios of disasters in order for people who are involved in emergency preparedness to do some exercise in case of any unforeseen disaster and with this software improve their ability to face real scenarios and be able to handle the situation in a systematic, controllable way. Each scenario has some sections and questions which provides discussion, problem solving and decision making for those who are involved in the exercise. Participants should play the assigned roles and positions and make step by step decisions based on the disaster scenario.

In each disaster, an outbreak of a disease is recognized by complaining of a person in the Emergency Department with similar symptoms of the disaster. For an instance in the Anthrax scenario a 55 year old male arrives in the emergency department of the hospital with symptoms of fatigue, fever, cough and chest pain. He started having short breath, blue nails and higher fever. Clinical tests confirmed Anthrax and more people are infected with the virus. The Software helps to handle the situation by asking questions about how and to who share information about the disaster, activation procedure and staffing plan, agency’s plan as well as federal agency’s plans. Mass Clinics are set up and different agencies are collaborating to handle the situation. The software is also a good exercise for those who are involved how to deal with issues such as mass panic, chaos erupts and media that interferes with hospital operations.

Solutions are also offered for each disaster and long term effects are also discussed. Long term effects are including medication effects, economic downturns, increasing unemployment and other social issues are also discussed. Also demobilization after the event, environmental surety of affected areas are also considered in the software.

Many jurisdictions are planning to setup and operate PODs out of convenient facilities such as schools. Patients come to the POD, and the staff there serve them. We have spent a lot of time developing our Clinic Planning Model Generator to model such PODs.

Karen McKinnis, at the Springfield-Greene County (Missouri) Health Department, has been developing and testing two other types of PODs. I thank her for sharing this information with me and allowing me to post it here.

The first type, a Private POD, uses a large room (like an auditorium) in which patients are instructed by a staff person who periodically gives instruction on how to complete the medical history form. After completing their forms, the patients leave through different color-coded doors that lead to different areas for dispensing different medications (like Tamiflu, Relenza, or NPIs). Families that need a combination of medications go another distinct dispensing area. The staff person repeats the talk periodically.  (I would note that this is an interesting variant of a traditional POD that uses a batch education, almost like making patients watch smallpox education video at a mass vaccination clinic, and can be modeled in a similar way.)

The second type, a Group POD, is much more radical.  In this case, the health department works with a business, a university, or some other organization with medical staff. The medical history forms are sent to the organization ahead of time. Their folks complete the forms and return them to the health department, who prepares the needed medication and sends it to the organization. The organization then dispenses the medication at their site by their medical staff.

When dispensing antibiotics, determining who should get which medication is an important question, but the rules can be confusing, and paper forms are not efficient.
Working closely with our partners at the Montgomery County, Maryland, Advanced Practice Center for Public Health Emergency Preparedness and Response, we have developed and released eMedCheck, an electronic medication screening form that can be run on a PDA.

The screening form determines who should receive Doxy, Cipro, or neither. The decision rules used in eMedCheck are identical to those in a medication screening form developed by the National Capital Region.

Instructions for downloading the software and installing it on your Palm PDA are included with the software.

How will global climate change affect public health departments? The National Association of County and City Health Officials (NACCHO), Environmental Defense Fund (EDF) and George Mason University have conducted a survey of local public health directors to answer this question. The report, published by EDF, shows that local public health directors worry that climate change will cause serious public health problems, believe that they are not ready to deal with the impacts of climate change that could occur, and desire additional resources to improve their preparedness. (Their most common concerns were heat waves, poor air quality, poor water quality, droughts, fires, and storms.) The report’s recommendations follow naturally from these beliefs. Included is the recommendation to “Develop improved climate and weather modeling capacity for local scale assessments.”

Interestingly, the report includes the following observation:

there are many opportunities for synergy between existing public health preparedness activities
(e.g. bioterrorism, pandemic flu, and all-hazard preparedness) and those addressing health threats from climate change.