A Potentially Fatal Illness, An Unpredictable Vaccine

To be protected from the achy, sneezy, feverish flu, people need to get a new formulation of the vaccine every year, for one thing. But few are so consistent. What’s more, manufacturers grow the influenza vaccine in chicken eggs, and no one can predict how much vaccine a given egg will produce. All of that uncertainty—on both the demand side and the supply side—provides a unique set of challenges for companies that want to both produce and profit from the influenza vaccine. In turn, those manufacturing concerns can cause trouble for consumers who want to stay healthy but may search in vain for doses of the vaccine, says Sarang Deo, an assistant professor of managerial economics and decision sciences at the Kellogg School of Management.

In a new paper, Deo and colleagues consider, for the first time, all of the factors that affect the quantity of flu vaccine produced and the number of people who seek it each year. Their results may help companies turn a bigger profit on the flu vaccine. But more important, the work suggests ways for public health agencies to improve their strategies for doling out the vaccine.

“This production uncertainty is very, very unique to the flu vaccine,” Deo says. “You don’t see that in many other drugs. You can plan to produce a certain number of doses, but the actual number of doses produced is random. At a basic level, companies don’t enter the market because they don’t think it’s profitable. If you combine these two things, you can get severe shortages of the vaccine.”

Roots of Inefficiency

It was the unpredictable nature of the flu that sparked Deo’s interest in seeking new ways of studying supply and demand for the vaccine. His first study on the subject, published in 2008 in the operations management journal M&SOM, used mathematical models to show how uncertainty in the production process can hamper profitability for manufacturers. And that, in turn, makes firms wary to enter the market.

To follow up on those findings, Deo worked with colleagues Kenan Arifoğlu, a doctoral candidate at Northwestern University, and Seyed Iravani, a professor of industrial engineering at Northwestern. The researchers created a new set of models to combine two main ways of thinking about the flu vaccine’s supply chain problem.

One school of thought focuses purely on the uncertainty inherent on the supply side, driven wholly by the unpredictability of chicken eggs. The other school of thought focuses purely on the erratic demand for the vaccine, which is driven by the unpredictability of human behavior. According to this behavior-based view, it would be best for the health of society as a whole if everyone got immunized. But since the social benefit of the vaccine is higher than the personal benefit for each individual, economists argue, fewer people get vaccinated than would be best for the entire population.

Together, the uncertainty of supply and the self-interested behavior of individuals create a lot of inefficiency in the distribution of the flu vaccine. But how much inefficiency is there? The answer, the new study found, is: It depends. Instead of adding up like a mathematical equation, the two sources of inefficiency actually interact with each other in complicated ways each year, the researchers report in a study that has been submitted to the journal Management Science. What’s more, demand-side inefficiency depends on supply uncertainty, and the supply side-inefficiency depends on the infectiousness of each year’s circulating strain.

Imagine, for example, a year in which manufacturers have produced fewer-than-expected doses of the influenza vaccine and the virus is particularly nasty, so that more people than usual are vulnerable to it, much like last year’s epidemic-inducing strain of H1N1. In cases like these, demand exceeds supply. Shortages follow. And the result, the new model predicts, is an ensuing cycle that makes the situation worse.

“We see that shortages are accentuated not only because the quantity is lower than normal but also because more people are desperate for it,” Deo says. “More people come out and search for the vaccine precisely because the quantities produced are lower. And everybody knows that if my chances of getting the vaccine are lower, that’s true for everyone and that means my chances of getting infected are higher.”

“If there’s a huge surplus, then the effect sort of reverses,” he adds. “People think: Oh, there’s a lot of vaccine, people will go and get it, so my incentive is that much less.”

From Mathematical Models to Public Health

By offering a nuanced and comprehensive way of thinking about the dynamics of flu-vaccine supply and demand, Deo’s models might change the way public health officials develop their yearly strategy for doling out the vaccine in a way that benefits the greatest number of people. The new work, for example, counters the advice of traditional economic theory, which argues that because people as a rule don’t get vaccinated as often as they should, policy-makers should do whatever they need to do to provide incentives for everyone to get the vaccine.

Instead, Deo says, the benefits of intervening on the supply side in a given year will depend on how much and what kind of demand there is for the vaccine. Likewise, the benefits of intervening on the demand side will depend on the year’s supply issues.

In a year where there is a lot of uncertainty in production and the disease is particularly infectious, for example, the most effective strategy for the Centers for Disease Control and Prevention (CDC) is to focus its energies on boosting supplies. In fact, the more infectious the disease becomes, the less effective it is for the CDC to work on influencing demand, because people will naturally seek out the vaccine at higher rates than normal.

The organization does its best to intervene in both supply and demand, but it doesn’t always strike the right balance. A more scientific-based strategy might help.

“We found using our model that if you don’t prioritize high-risk groups when there is higher demand, there is a lower chance that people with the highest priority will get vaccinated,” Deo says. “People focus on how many people get vaccinated. But our model shows that it also matters who gets vaccinated.”

Deo’s work could also help policy-makers and public health officials decide, in any given year, whether they need to focus their efforts on what consumers are doing or on what vaccine manufacturers are doing. “What our model really gave us,” Deo says, “was a set of insights which said that the effectiveness of intervening on the demand side depends on how big the problem is on the supply side, and vice versa.”

If influential organizations and regulatory bodies catch on, his work may lead to healthier, less achy winters and fewer sick days on the couch for a bigger chunk of the population. “I would like to strike a dialog with the CDC,” Deo says, “and maybe influence their thought process.”

Related reading on Kellogg Insight

Rational Rationing: Allocating scarce resources in uncertain conditions

Fetal Influence on Familiar Ailments: The effect of in-utero conditions on long term health