Preventative medicine: Calculating disease burden and fighting flu with computer models

August 02, 2011

Anyone who’s ever played SimCity or Civilization knows there’s a whole genre of computer games that let players make public policy decisions and watch the likely outcomes on screen. In such games, simulated populations suffer or thrive based on the gamer’s choice of government action.

Wouldn’t it be great if real policymakers could see computer simulations of their decisions before making them? In Arizona’s Maricopa County, with a population of 4 million residents, the Department of Public Health (MCDPH) will soon have such a computerized tool to use in fighting pandemic flu.

Building on an eight-year history of applying academic research to practice-related issues, a team of scholars from the W. P. Carey School of Business have been developing a computerized model designed to calculate  influenza’s societal costs based on changeable factors, such as how early vaccines get distributed and who gets the first doses. The tool draws upon past engagements in which scholars were contacted by MCDPH to address health issues through technology and business acumen. This work will help cost-justify flu-fighting expenditures and, most likely, help public health professionalsthwart flu’s spread, as well.

Dollars and disease

The tool’s development effort starts with the notion of “disease burden,” which Maricopa County Department of Public Health physician Rebecca Sunenshine explains is “the impact of any health problem in a specific area, such as financial costs, child care for kids who can’t go to daycare and level of sickness or what we call morbidity.” Lost wages, lost productivity, the costs of treating the illness, emergency-room visits and hospitalizations – these all come into play.

So does the cost of death. “When we look at the cost of death, we look at what’s lost in human capital,” explains 

Raghu Santanam, professor of information systems at the W. P. Carey School of Business and one of the researchers working on the decision-making tool. “You’re looking at productivity loss if a child doesn’t grow up and contribute to society. What’s potential value of a person over a lifetime?”

"If you have more possible productive lifetime ahead of you, death is a bigger potential loss,” adds W.P. Carey School professor Ajay Vinze, who is leading the research team on the project. “It’s a gross simplification, but it’s the best you can do.”

Calculating disease burden is a large part of the deliverable the Vinze and Santanam will be handing over to Maricopa County this month. In order to come up with a mathematical model for this calculation, the team had to first scour existing research to see what costs are involved and their likely amounts.

Also included in the model will be the costs involved in preventing disease, something that can be done easily with pandemic influenza through vaccination. Plus, the model has to factor in demographics of those who get vaccinated, as medical researchers have found it’s more cost-effective to vaccinate some populations over others.

Prioritizing prevention

 “One of the main things we’re looking at is the advantage of doing school-based vaccinations,” says Dr. Sunenshine. “That would target school children, whom we know are, according to the literature, in the age group that spreads influenza the best.”

Another factor involved in the decision-making is “herd immunity,” the notion that “if you get to a certain level of vaccination in a community, the disease doesn’t have any place to spread,” Sunenshine continues. According to her, medical experts think that the herd-immunity vaccination level may be in the neighborhood of 80 percent, and she points to the success of vaccinations for childhood diseases, such as mumps and rubella.

"Because of herd immunity and the germ-spreading power of school kids, Maricopa County is interested in “evaluating whether vaccinating all of the school-age children in the county will prevent spread of disease in the entire population, including the elderly,” Sunenshine adds.

Pricing good policy

Sunenshine and others in public health are interested in cost-benefit analysis of vaccination strategies because of simple economics. “We’re in tight financial times right now, and federal funding to purchase for certain groups is going away,” she says. And, not all insurance pays for vaccination. “If we had data to show that vaccinating was a benefit to an insurance company, we might be able to share it and convince more insurance companies to pay for vaccine up front.”

Legislators would be an audience for the data, as well, according to the W.P. Carey School’s Ajay Vinze. He says public health officials advise legislators and regulators on the “impact of a government body committing funds to different diseases. If we put our money into vaccination, how does that investment come back? It’s a cost-benefit analysis.”

Ultimately, such a cost-benefit analysis is what will come back to physicians and other officials at the country health department. Getting to the figures is complicated, in part because the mathematical model must identify and factor in the various costs associated with disease burden, and also because you need to know how disease will propagate to fully calculate its impact.

Simulated cities

As is turns out, the W. P. Carey School had already created a model to simulate the spread of flu in the Phoenix-Southern California region. “We created a simulated population based on census figures and other data,” explains the school’s Santanam.

For instance the team used Bureau of Transportation and census county to county worker flow data for travel patterns to see likelihood of Phoenix-L.A. travel. “We built a tool where you could input how fast the flu was spreading, what populations might be affected in different regions and various responses to the pandemic,” he continues. “We looked at vaccination, but also at non-pharmaceutical responses, such as closing transportation between regions or closing schools and offices.”

The researchers used this previous simulation work for the current project and also added three other Arizona counties to the mix, Vinze explains. “What our work does is take the raw numbers of vaccination and outbreak rates and then project the actual outcome of the disease,” he continues. “Then you take those numbers and feed them into the spreadsheet to show the cost benefit.”

Such a view of disease is “not typical for public health officials,” Vinze says. “They’re in the business of saving lives.”

But, budgetary constraints are pushing health officials to take new approaches to service delivery. “They have to justify why they’re asking for money to vaccinate or do things proactively.” This feeds into the interest in herd immunity.

“Until the H1N1 swine flu incident in 2009, there was no nationwide tracking of vaccinations,” Santanam says. “Now the Centers for Disease Control is tracking across regions. They’re finding out vaccination rates by age group. Such data can help better understand how herd immunity can save lives and money for stakeholders.”

Preliminary data from the model Vinze and Santanam have built indicate that the costs of influenza are high and early vaccination, if given to the right populations, could save society plenty. In looking at the effects of pandemic influenza in Maricopa County, Vinze summarizes impressive findings: “If you don’t vaccinate in the county, the disease can can be extremely costly to the community.” Taking the right vaccination approach, he adds, “could cut costs down significantly.”

He adds: “Despite the obvious benefits from efforts of the department of public health, the department is not immune from having to justify return from efforts underway. The cost-benefit evaluation our team has provided is in line with expectations of these tough economic times.”

Bottom Line

  • Public health officials in Maricopa County, Arizona, have tapped researchers at the W.P. Carey School of Business for help evaluating the “disease burden” of pandemic influenza.
  • Disease burden is the impact of a disease, such as costs for treatment, hospitalizations, lost productivity, and even the lost potential productivity resulting from untimely death.
  • The W. P. Carey School’s research has produced a computerized model to calculate disease burden and cost benefits of preventative measures, such as early vaccine.
  • Through the model, Maricopa County will be able to demonstrate the cost-effectiveness of prevention and treatment efforts.

-- Photo credit Flickr, creative commons, ZaldyImg's photostream