News

Laura J. Kornish
Jonathan B. Wiener

This week, on Nov. 11-12, the World Health Organization will convene a global summit to prepare for the next influenza pandemic. At the same time, Americans are facing the flu season with only half the anticipated supply of vaccines. Next year, we need a more diversified supply. But even if the supply could be assured, a frightening question remains: will the vaccine actually work?

Like a shell game in which the contestant has to guess which shell hides the prize, every spring the Food and Drug Administration has to guess which strains of the virus will be dominant the next fall. The circulating strains of the influenza virus change every year. Because it takes several months to manufacture the vaccines, the FDA chooses the virus strains to put in the vaccine long before knowing which strains will actually strike. So even if we had the full supply, and even with the best available forecasting by FDA, there is a real chance that vaccine could have been formulated for the wrong strains and would do little to protect us.

The 1918 Spanish flu pandemic killed over 500,000 Americans and more than 20 million people worldwide (out of a smaller total population than today). Pandemic flu returned in 1957 and 1968; the next crisis may be imminent. In recent years, with vaccines, there have been about 36,000 deaths due to the flu in the U.S. each year.

But every year we risk choosing the wrong shell and winding up with an ineffective vaccine. We can do better. Public health can be better protected by improving the decision-making process and reforming the incentives in the flu vaccine system.

First, FDA could adopt criteria for deciding when to wait longer to learn more before selecting the vaccine formula. Even though delays reduce the time available for vaccine production, the agency should take seriously the option to defer the recommendation to gather more information about the circulating strains. Clearly, there is a tradeoff between knowing more and producing more. A perfect vaccine that arrives too late is no good; but neither is a punctual vaccine that targets the wrong germs. With a new, potentially highly contagious virus strain, a small delay in the decision may make a crucial difference in confirming or disproving the rapid spread and hence the need to include it in the vaccine.

Second, FDA could make its vaccine decision in stages. Currently, the FDA makes a recommendation each year for a trivalent (three-strain) vaccine, all at once. If FDA decides on each strain as information warrants, vaccine manufacturers have more flexibility. Starting earlier on the first strain gives them precious extra time for production of the later strains, essentially increasing their capacity.

Third, FDA should consider adding a fourth strain to the trivalent formula. The current influenza vaccine licenses are all for trivalent formulations, but three is an arbitrary number. While four-strain vaccines would require additional testing, adding the fourth strain would enable the vaccine to protect against a wider variety of influenzae.

Fourth, the pharmaceutical companies can play a role. While the government chooses the formula for the influenza vaccine, the production and distribution are largely private tasks. Instead of waiting to start production until the FDA decides on the three-strain formula, the manufacturers could start producing one or two of the uncontroversial strains sooner. This move would free up production capacity in the subsequent months, allowing for a higher overall quantity of the complete vaccines. There is a cost -- the possibility of wasted production -- but the gamble may be worth it financially — and for public health. The government could even help insure the manufacturers against such losses.

More generally, the government should evaluate ways to make the influenza vaccine market more financially attractive to producers — and to shorten the time needed for production. Shorter production time would give the FDA more time to assess which strains will infect us, and thus help FDA formulate a much more effective vaccine. FDA should consider ways to reward efforts to produce vaccines more quickly. One promising option is cell culture, which may prove faster, cheaper and safer than the current chicken egg culture method.

This year’s supply shortfall should be a wake-up call to help improve the influenza vaccine formulation and production system. The typical reaction to a shell game is to walk away from the table. But we can’t afford to let a flu pandemic strike us unprepared. We need to change the game, giving us better information and more time to get it right.

Laura Kornish is Assistant Professor and teaches decision sciences at The Fuqua School of Business at Duke, and Jonathan Wiener is the William R. & Thomas L. Perkins Professor and teaches risk policy at Duke Law School.