Stanford engineer helps curb disease and guard against bioterrorism
The high-stakes global logistical challenges of the Second World War gave birth to a new form of engineering focused on optimizing the manufacture, deployment and use of personnel and material to support U.S. military efforts across the globe. In the post-war years, the field that came to be known as operations research transformed factories and businesses as they churned out more, better and less expensive goods ranging from cars to computer chips.
Today, operations research is part of the broader field known as management science and engineering, but the principles remain: efficiency and optimization in all things. Margaret Brandeau, the Coleman F. Fung Professor at the Stanford School of Engineering, is taking operations research in a new direction, applying the rigorous modeling and mathematical approaches of operations research to challenging problems in public health.
“We’ve found that the complex systems models that have been so successful in streamlining operations and manufacturing are equally adept at helping us determine the most efficient course of action in public health,” Brandeau says.
Brandeau’s work has ranged widely, from improving how resources are invested in the fight against the spread of HIV, to answering governmental policy questions that resulted in the mass vaccination of millions of children in China against hepatitis B. She has even been called upon to model how the country should plan a response to a bioterrorism attack, should one ever occur.
In a perfect world, all public health decisions would be driven by data from randomized clinical trials, but such trials can be costly, impractical or, in some cases, unethical. Modeling fills the void, helping to evaluate the likely results of various public health programs without the profound outlays of resources or ethical questions.
“People in public health often say, ‘We can’t make a decision because we don’t have enough information.’ But what they don’t see is that doing nothing is also a decision,” Brandeau says, adding, “An informed estimate is far better than none at all.”
With its emphasis on epidemiology and the spread of disease, public health is no stranger to modeling. But Brandeau says that management science and engineering brings a more nuanced way to answer some tough and politically challenging questions by building models to help solve difficult problems that have stymied policy makers in the past.
“Operations research was founded precisely on the idea of trying to solve large, intractable problems with multiple and often unknown variables,” she said.
Beating a Silent Killer
Brandeau recently worked with the Asian Liver Center at Stanford to explore different approaches to combatting hepatitis B infection in Asian populations.
Hepatitis B is particularly acute among the Asian community. While just one in a thousand non-Asian Americans has hepatitis B, in the Asian community worldwide the figure is as high as one-in-ten. Infection usually happens in childhood, but the virus can lie dormant for decades only to emerge later in life in the form of lethal cirrhosis or liver cancer. Yet hepatitis B is easily preventable with an inexpensive vaccine. In addition, new antiviral treatments can control the virus in infected people, allowing them to live normal lives.
Determining how best to control hepatitis B is a problem with many alternatives and a high potential impact—exactly the sort of challenge ideally suited to operations research.
“We’re looking to provide one-sentence answers to the cost-benefit question: with limited funds, what’s the best way to have the biggest impact?” Brandeau says.
In the U.S., Brandeau worked with the Asian Liver Center and the Centers for Disease Control and Prevention (CDC) to study the best way to control hepatitis B among Asian and Pacific Islanders, who have a far higher prevalence of infection than the general population. Her analyses helped reshape U.S. public health policy. Her models showed that it would be cost effective to screen adult Asian and Pacific Islanders for infection and subsequent treatment and to vaccinate anyone in close contact with the infected people.
In China, where many children have not been vaccinated against hepatitis B, Brandeau evaluated the likely costs and benefits of providing “catch-up” vaccination to children who were not vaccinated at birth. Brandeau’s recommendations have led to a dramatic shift in policy in China, which now provides free catch-up vaccination to all Chinese children under the age of 15—as many as 140 million children. While the effort could cost half a billion dollars, her study shows that it will prevent nearly 8 million acute infections, 400,000 chronic infections and almost 70,000 deaths from hepatitis B, and will save $1.4 billion in future health care costs.
Outsmarting HIV
With HIV prevention, Brandeau faced a different sort of challenge, one that is part financial, part political. According to the Joint United Nations Programme on HIV/AIDS (UNAIDS), worldwide an astounding 6,300 new HIV infections occur every day.
In one study, she modeled a number of HIV prevention strategies in the US, including the daily use of the antiretroviral drug known as Truvada by men who have sex with men who are at high risk of contracting the virus. Truvada has been shown to prevent the chance of HIV infection by as much as 70 percent, though it comes at heavy cost, $27 per day, per person.
Despite this, Brandeau’s analysis showed that Truvada could be a cost-effective HIV prevention strategy, but only if administered to the 20 percent of people at highest risk of infection. Her analysis, sponsored by the National Institute on Drug Abuse, showed that such a daily regimen in the United States would be cost effective and would reduce new HIV infections by an estimated 13%. Beyond that, however, there proved to be diminishing returns between prevention and cost.
She has also evaluated various HIV control strategies in areas of high HIV infection risk such as Russia and in the injection drug user communities of Ukraine. In the future, Brandeau will be applying her models to look at co-infection that happens often between HIV and other diseases like tuberculosis, malaria and hepatitis C.
“Margaret was one of the first people to introduce me in the mid-1990s to the power of modeling healthcare issues,” said Eduard Beck, Regional Strategic Information Adviser for UNAIDS. “I find her attitude extremely refreshing and inspiring, especially as she is very open to discussing and thinking about the big policy issues as well as the practical implications with stakeholders, ranging from members of civil society to policy makers to fellow modelers: a rare gift.”
Improving Bioterrorism Readiness
With the anthrax attacks in 2001 that killed five people and infected 17 others nationwide, suddenly bioterrorism became a more real threat than ever. Brandeau was soon applying her modeling techniques to preparedness strategies.
Among the questions she and her collaborators explored included an investigation of the supply chain for antibiotics that would be necessary to combat a large-scale anthrax attack. How much antibiotics would be needed? Where should the antibiotics be stored? How rapidly could they be dispensed to those in need? Where would the bottlenecks in the supply chain be?
“We learned that the supply of antibiotics is not the problem; it’s the dispensing of the drugs that will limit our response to an anthrax attack. Rather than investing in supplies of antibiotics, cities need to develop improved plans for dispensing the antibiotics,” Brandeau says.
Brandeau’s models showed that stockpiles should be amassed in local hospitals but only in the areas of greatest risk—metropolitan communities with high population density where an attack would likely prove most lethal.
Her models went so far as to question whether putting emergency doses in every household in the U.S. would be a wise investment. She found that such a mammoth effort would be financially and practically prohibitive.
Next, Brandeau was tapped by the U.S. Department of Health and Human Services to sit on a scientific advisory board that oversees the CDC’s preparedness efforts. This includes the $5 billion U.S. Strategic National Stockpile.
“This is a war chest for any public health emergency—a flood, hurricane, anthrax attack and so forth. But we lacked a basic understanding of what exactly the stockpile should be as our nation’s needs evolve. Modeling can help in planning the contents of the stockpile,” Brandeau says.
Modeling Good Decisions
As the breadth of Brandeau’s work proves, public health and engineering have much in common and the possibilities for applying the sophisticated modeling techniques of operations research are manifold.
“The goal of operations research modeling of public health strategies is not absolute certainty and precision, but instead a way to clarify what good decisions look like,” Brandeau says. “What’s needed is to derive answers that policy makers can use to make better policy decisions—and thereby make the best possible use of our limited public health resources.”
