Study Projects Flu Pandemic Could Kill 81 Million Worldwide
LONDON – A flu virus as deadly as the one that caused the 1918 Spanish flu could kill as many as 81 million worldwide if it struck today, a new study estimates.
By applying historical death rates to modern population data, the researchers calculated a death toll of 51 million to 81 million, with a median estimate of 62 million.
That's surprisingly high, said lead researcher Chris Murray of Harvard University. He did the analysis, in part, because he thought prior claims of 50 million deaths were wildly inflated.
"We expected to end up with a number between 15 and 20 million," Murray said. "It turns out we were wrong."
The new work is published in Saturday's issue of the journal The Lancet.
The 1918 flu outbreak killed at least 40 million people worldwide. But flu pandemics have varied widely in their severity. The most recent, in 1957 and 1968, were relatively mild, killing 2 million and 1 million people worldwide respectively.
To get their estimates, Murray and his colleagues examined all available death registration data from 1914 to 1923. There was sufficient information from 27 countries, including numbers from 24 U.S. states and nine provinces in India.
The researchers compared death rates during the pandemic to average death rates before and after. That revealed how much the pandemic flu contributed to death rates, a figure called excess mortality. They then applied the excess mortality data to worldwide population data from 2004.
If their median estimate of 62 million flu deaths occurred in a single year, the total number of deaths from all causes worldwide would more than double, jumping by 114 percent.
One surprise in the new study was the huge variation in how different countries would be affected by a pandemic. The study estimates that 96 percent of the deaths would occur in the developing world. Murray and colleagues noted there was a 30-fold or more variation in mortality.
"That tells us it's not just the genetic makeup of the virus that will cause deaths, but that there are a lot of other things that intervene," he said.
Determining the mitigating factors might help avert a catastrophe. "If we can answer that question, we may unlock the mysteries behind which non-pharmaceutical strategies could significantly decrease mortality," said Murray.
Population density, nutrition and immune status could all play roles, he suggests.
"We know that even if we have much lower numbers of deaths worldwide than in 1918, the world will be severely stressed," said Dr. Keiji Fukuda, coordinator of the World Health Organization's Global Influenza Program. "Speculating about the possible numbers is an interesting exercise, but the really important thing is, what do we do about it?"
Since the pandemic threat rose, with circulation of the H5N1 bird flu virus on a large scale in late 2003, the global community has bolstered its pandemic preparedness plans. Medical systems today are far stronger than they were last century, and the availability of antivirals and antibiotics — which did not exist in 1918 — should help greatly. Still, many of these advances remain out of reach for poor countries.
Another question is the impact a flu pandemic would have on those infected with HIV. Seasonal influenza exacts a heavy toll on those with weakened immune systems. So, in the case of a new pandemic flu, Murray's estimate might be optimistic.
And while the Spanish flu often has been regarded as a worst-case scenario, there is no guarantee the next pandemic will not be even more deadly. Despite the tens of millions of deaths the 1918 flu caused, the death rate among those infected was approximately 2 percent. The fatality rate for the H5N1 virus is about 60 percent.
However, experts think that if H5N1 were to evolve into a strain easily transmissible between people, it would also become less deadly.
"It's not in a virus' interest to kill its hosts so readily, otherwise it can't reproduce," said Dr. Ian Gust, a flu expert at the University of Melbourne, Australia.
Still, there is no guarantee that H5N1 would become less deadly.
If it doesn't, "we would be in for a devastating impact," said Gust. "All bets would be off."