Epidemiology and economics merged in a groundbreaking analysis of West Nile Virus data by SMU alumnus Robert Ware Haley ’67, M.D., and Tom Fomby, SMU Professor of Economics.
The interdisciplinary University-community research collaboration reveals several key findings about West Nile outbreaks and points to the use of a mosquito vector index rating system to trigger early intervention. Those results are published in the July 17 issue of JAMA (The Journal of the American Medical Association), the prestigious peer-reviewed medical journal.
Haley, Chief of Epidemiology and Professor of Internal Medicine at UT Southwestern Medical Center, and Fomby, Director of the Richard B. Johnson Center for Economic Studies in Dedman College, joined forces with Wendy Chung, Chief Epidemiologist for the Dallas County Health and Human Services Department, and her colleagues, Christen Buseman, Sibeso Joyner and Sonya Hughes, in the study. James Luby, M.D., Professor of Internal Medicine at UT Southwestern and a longtime research collaborator with Haley, is also an author.
>Read more about their West Nile research
A senior author of the journal article, Haley calls the research a “stone soup” project, referring to the folk tale that demonstrates how cooperative efforts benefit the entire community. “Everyone contributed their data and expertise to produce significant advances in our understanding of West Nile.”
The perfect storm of conditions that created the 2012 public health emergency in Dallas County presented an unprecedented opportunity to study the anatomy of the nation’s largest West Nile outbreak, says Haley, who has been involved in research on mosquito-borne illnesses since he was a medical student at UT Southwestern in the 1960s. He spent 10 years with the Centers for Disease Control and Prevention, serving as a commissioned officer in the U.S. Public Health Service, before returning to the medical center in 1983 to found the Division of Epidemiology and Preventive Medicine.
Last year, an estimated 400 people in Dallas County suffered mild to severe West Nile infections. The county’s 20th death related to that outbreak was recorded in April 2013.
The econometric approaches Fomby brought to the statistical analysis of the unusual data were crucial, Haley says. This latest research cross-pollination follows previous work – along with Wayne Woodward, Professor and Chair of SMU’s Department of Statistical Science – on two noteworthy appendicitis-related studies published in 2010 and 2011.
“With the publication of these two papers we saw how techniques developed in one field can be usefully applied in other fields when researchers embrace collaborators outside of their own areas of expertise,” says Fomby.
Fomby’s specialized knowledge of count time series models exposed the value of the mosquito vector index as a leading indicator of subsequent West Nile Virus outcomes. And, his use of event study analysis – which he says is “fairly unique to applications in economics and finance” – showed aerial insecticide spraying was not associated with increases in hospital emergency room visits for respiratory symptoms or skin rashes.
Another key discovery materialized when Haley combined data collected by the county health department – including patient statistics and mosquito-trap test sample results – with weather information.
“Major outbreaks of West Nile occurred in 2006 and 2012. Both of those years had the fewest hard-freeze days in the winter and, overall, warmer than average temperatures,” he says.
Merging health department data with census tracks located another marker: areas of higher property values, higher housing density and higher percentages of unoccupied homes are at higher risk. In Dallas County, the data showed West Nile clustering in the Park Cities and North Dallas, areas with environments ripe for house mosquitoes, which are more likely to transmit the disease.
Along with the findings, the researchers provide an instruction manual for health officials in other counties to calculate their vector index by plugging in their own data.
“Virtually every community in the country has the potential for a West Nile outbreak, and provided with this prediction model, they can conduct their own analysis and determine when to act,” says Haley.
Both Haley and Fomby say they look forward to continuing a partnership that stems from a Town and Gown Club at SMU meeting in 2006.
“I gave a talk on how data mining (also called big data) was affecting many areas of our lives on a daily basis,” Fomby recalls. “Robert commented on how he saw quantitative reasoning and data analytics significantly affecting his fields in the future.”
Haley, honored as a Dedman College Distinguished Graduate in 2008, traces his SMU roots to its early days. His maternal grandfather, Samuel D. Ware, was a strong proponent of building a Methodist University in Dallas. His parents, Arvel E. Haley and Charlotte Ware Haley, met in a music class at SMU. In fact, more then 20 members of his extended family are SMU alumni.
After finishing pre-med requirements, he also completed a bachelor’s degree in philosophy and social sciences at SMU and taught a philosophy survey course for a year after graduation. That liberal arts background continues to inform his work.
“The hardest part of research is asking the right question,” he says, “and there’s nothing like studying philosophy, in some depth, to understand the right way to ask a question to get the answer you need.”
In addition to research projects with Fomby, Haley has worked with Woodward and Richard Gunst, Professor of Statistical Science, and William Schucany, Professor Emeritus, on pioneering brain imaging data studies of veterans with Gulf War illness. Haley holds the U.S. Armed Forces Veterans Distinguished Chair for Medical Research Honoring America’s Gulf War Veterans at UT Southwestern.
“When I had a really difficult problem in blazing new ground in statistics, I was lucky enough to have one of the top departments here in my backyard,” says Haley. “What they did was absolutely original, creative and brilliant.”
– Patricia Ward
Categories