Web Exclusives: Diseases
Disease Modeling Efforts Gain New Ground
To help the nation—and the world—understand and prepare for contagious outbreaks, new research groups join the Models of Infectious Disease Agent Study (MIDAS). Their work will simulate disease spread, evaluate different intervention strategies and help inform public health officials and policymakers.
Here's a quick look at each project:
Diane Lauderdale, University of Chicago
Charles Macal, Argonne National Laboratory
MRSA, or methicillin-resistant Staphylococcus aureus, can cause dangerous skin and other infections that no longer respond to many of the antibiotics that doctors routinely use. MRSA has spread rapidly during the past decade and last year killed more Americans than AIDS.
To get a handle on how MRSA spreads, epidemiologist Diane Lauderdale of the University of Chicago and systems engineer Charles Macal of the Argonne National Laboratory will take advantage of a computational technique called agent-based modeling. Lauderdale likens the approach to computer games that simulate the response of an urban population to various threats, but says the modeling is more complex and is based on real data.
Lauderdale and Macal's model will include data on where, when and how people move around. This will help them understand how patterns of contact and behavior among individuals affect MRSA spread. The researchers will also use the model to test the effectiveness of different interventions, such as public health informational campaigns or enhanced hygiene in institutional settings, like schools or work places.
"Epidemiologists have used models to simulate the spread of infectious diseases for some time, but earlier versions were much simpler, which limited their value in guiding public health response," says Lauderdale.
Macal adds: "The more sophisticated models being developed today are possible because of the enormous advances in computer and software technology in recent years."
Lauderdale, Macal and their research team bring substantial expertise on the biology of MRSA to the project and their models will provide an added tool with which to combat the growing threat of the infection in the U.S.
Ira Longini and Elizabeth Halloran, Fred Hutchinson Cancer Research Center and University of Washington in Seattle
Almost everyone can remember a time when they had to make a complex decision quickly and relied on their instincts in deciding what to do. Public health officials are put in situations like these every time a contagious disease emerges in their area. Without enough time or information to make a fully informed decision, they must often choose interventions and allocate resources based on little more than guesswork.
To take some of the uncertainty out of responding to infectious disease outbreaks, biostatisticians Ira Longini and Elizabeth Halloran and their colleagues at the Fred Hutchinson Cancer Research Center and the University of Washington in Seattle are developing computer models that simulate outbreaks. When a real outbreak occurs, they and others can enter specific information into the model, like contagiousness and the population density at the source of the outbreak, and quickly identify interventions that could stymie the spread of infection.
"People have been building models of infectious disease spread for over a hundred years," says Longini. "But today's models are far more useful because of advances in computing and integrating field and scientific data. In recent decades, modeling helped control HIV in the United States and foot and mouth disease in the United Kingdom."
Longini and Halloran are modeling some of the world's major threats, like flu, cholera, dengue fever and tuberculosis, drawing on their expertise in biostatistical and epidemiological approaches and their experience in carrying out field trials. Their work in the field, such as a recent trial in Senegal to test the effectiveness of flu vaccination, gives them a unique real-world perspective that represents a valuable asset in building robust computational models.
Longini expects his models to be an effective tool in helping us prepare for bioterrorist and naturally occurring disease outbreaks and decide how to distribute resources in response to such events.
Donald Burke, University of Pittsburgh
Donald Burke, an international health expert at the University of Pittsburgh, sometimes thinks of his computer as a miniature version of a country or a continent, with millions of people walking around inside. In this virtual world, Burke can think through and test interventions computationally—"in silico"—before an actual outbreak occurs.
Burke's team uses real-world data to create artificial communities in which individuals are assigned to schools, workplaces and households. Once they introduce the virtual infection, they can track its spread through these simulated worlds. Burke's models have already been used by the World Health Organization in helping guide decisions on how much influenza antiviral medicine to stockpile worldwide.
The trajectory of an infection through a population depends on a dizzying array factors, and constructing a useful model involves taking them all into account. Herein lies Burke's strength.
His team includes epidemiologists, psychologists, microbiologists, behavioral scientists and public health practitioners. One group is developing computational methods to contend with the vast amounts of data that go into the model, while another is creating outreach and training programs aimed at educating a diverse group of scientists in modeling techniques.
Marc Lipsitch, Harvard School of Public Health
When an infectious disease first appears, public health officials often have just a trickle of information to inform decisions that could greatly impact the health of the community. Marc Lipsitch of the Harvard School of Public Health experienced this first hand when the U.S. Centers for Disease Control and Prevention asked him to help analyze incoming data in the early days of the recent H1N1 pandemic. His expertise in modeling the emergence of infectious diseases and how best to distribute antiviral medicines in the face of drug resistance were valuable in preparedness planning.
Lipsitch wants to make modeling resources, experts and methods widely available to public health decision makers. His team is developing epidemic "dashboard" software to give policymakers visualization tools to help them interpret trends and analyze data in the midst of an emerging epidemic. Early miscalculations of the fatality rate of SARS in the 2003 outbreak might have been avoided if better tracking tools had been available, says Lipsitch.
Lipsitch's team represents a diverse network of local collaborators, forming a nucleus for cross-disciplinary research in the Boston area. His group also plans to reach out to experts worldwide and to continue partnering with public health officials in Hong Kong, the Netherlands and the United States in developing training programs, software and research projects.
Alison Galvani, Yale University
Lauren Ancel Meyers, University of Texas at Austin
How would you react to news of an infectious disease outbreak? How much would you pay for a vaccine or antiviral medication? Who would you rely on for the latest information? Your answers—and those of thousands of others—are at the heart of this new MIDAS project led by Alison Galvani at Yale University and Lauren Ancel Meyers at the University of Texas at Austin.
Key to the researchers' modeling work is surveying people on how they perceive health risks. The researchers will use this information to build a dynamic model that simulates how changes in decision-making influence patterns of disease spread. The model will help them and others identify the strategies that improve adherence to interventions, optimize containment measures and ultimately reduce disease spread. Right now, Galvani and Meyers are using traditional surveys and even Facebook to ask people about H1N1, or "swine" flu.
Meyers says that her group is interested in tracking how people's answers change as public health officials issue new information or guidance about H1N1. The differences could reflect what happens to people's perceptions, behaviors and choices as a disease outbreak evolves. Incorporating this information, adds Meyers, will truly advance the field of disease modeling.