Who Died From Flu?

To the Editors:

Helen Epstein’s article “Flu Warning: Beware the Drug Companies!” [NYR, May 12] contains a significant factual error. It states that worldwide the death toll from the 2009 H1N1 influenza pandemic was 18,000. In fact, as the World Health Organization has tried, but apparently failed, to make clear, that number represents only laboratory-confirmed cases, and these numbers are a small fraction—possibly a tiny fraction—of the true death toll.

In the United States, there were 2,117 laboratory-confirmed deaths, yet the Centers for Disease Control estimate actual deaths in the US alone at between 8,870 and 18,300. Applying similar multiples to laboratory-confirmed cases around the world would yield 72,000 to 162,000 deaths. But that number, since it is still based on confirmed cases, still understates the deaths by a wide margin. Few countries—and none in the developing world—devoted much resources to testing cases. In many countries there were so few laboratory-confirmed cases that the number is meaningless. All of Africa had only 168 laboratory-confirmed deaths.

In addition, since the people who died were much younger than is normally the case from influenza, in terms of years of life lost the H1N1 pandemic was significantly more lethal than the raw numbers suggest.

Was H1N1 a phony pandemic? It certainly did not match the standards of past pandemics either in mortality or morbidity; in fact, there is a greater gap between 2009 and past pandemics in morbidity than in mortality. The WHO did not, however, have the benefit of hindsight. But it did try to learn from history. Just as in 2009, 1918 saw a mild spring wave of influenza. Indeed, it was so mild that in the British Grand Fleet patrolling the coast of Europe, 10,313 sailors were ill enough to report to sick bay but only four died. That fall the virus then became virulent. (In fact, the virulence was so different that a minority of virologists think there may have been two pandemic viruses in 1918; this is unlikely, especially given the fact that exposure to the spring wave provided as much protection against the lethal fall wave as a good modern vaccine.) In the spring of 2009 the WHO had no way of knowing to a reasonable level of confidence whether the virus would remain mild, as it did, or become lethal, as happened in 1918.

The WHO did the right thing in declaring the pandemic. We simply got lucky.

John M. Barry
Distinguished Scholar
Center for Bioenvironmental Research
Tulane University
New Orleans, Louisiana

Helen Epstein replies:

How many people actually died from H1N1 influenza during 2009 and 2010? I cite laboratory-confirmed deaths—meaning a cheap and simple blood test for H1N1 was carried out—and Professor Barry cites the Centers for Disease Control’s estimates based on mathematical modeling studies. However, some experts maintain that the CDC’s estimates studies overestimate influenza mortality, particularly among children.¹

It’s true that the death rate from H1N1 was relatively high among children. However, the total number of confirmed US pediatric H1N1 deaths was 371, a small fraction of the roughly 45,000 total annual deaths among US children.² While the death of any child is tragic, many of those H1N1 casualties had underlying conditions such as cancer, and would have been highly vulnerable to any infection.

The CDC estimates that the actual number of pediatric H1N1 deaths could have been 1,271 or even higher,³ but this is debatable, because any hospitalized child with severe influenza symptoms (easily identified by any doctor) would have been tested for H1N1, according to CDC guidelines at the time. Those with confirmed H1N1 were used as the basis for the CDC model, which then assumed that some cases would have been missed either because the children didn’t die in a hospital, or they tested negative when they actually had H1N1, or were missed for some other reason. The CDC model then comes up with a number based on these assumptions. But deaths in children from infectious diseases are rare in the US, and even those who didn’t die in hospitals would almost certainly have been autopsied (and tested for H1N1), especially given the intense “swine flu” panic back in 2009/2010. Also, the test is accurate and would have missed few cases. Because it’s unlikely that large numbers of actual cases of US child deaths from H1N1 were missed, the lab-confirmed count (371) is probably much closer to the true number than the modeled numbers Barry cites—which are in any case impossible to verify.

Moreover, as Barry notes, the CDC itself maintains that the 2009 H1N1 virus killed fewer people than ordinary seasonal influenza has in recent years, let alone the “pandemic”—(meaning novel)—viruses of 1918, 1957, and 1968.

Professor Barry’s suggestion that H1N1 mortality could have been higher than recognized in Africa seems unlikely. There, as elsewhere, influenza surveillance was much more thorough than in previous years. I happened to fly into Ethiopia and Uganda in early 2010 when the World Health Organization’s pandemic emergency declaration was in effect. A phalanx of nurses interviewed all arrivals about flu symptoms before they even reached the immigration desks, and the medical stores in both countries were stocked with Tamiflu that was never used. It’s a shame those nurses weren’t tending to the hundreds of Ugandan and Ethiopian children who die daily from malaria, malnutrition, and other diseases of poverty.

Professor Barry maintains that the WHO’s decision to declare an H1N1 pandemic emergency in the spring of 2009 was sound, even though the virus was known to be mild at the time, because the 1918 influenza pandemic may have occurred in two waves, a mild one in the spring and a far more deadly one in the fall. However, this isn’t a common pattern in the epidemiology of influenza.⁴ In the event, the June 2009 pandemic announcement caused canceled flights, closed schools, delayed conferences and business meetings, and overburdened health care systems all over the world. Why trigger such panic just because a mild virus might become more aggressive in four months’ time—assuming the “two-wave” theory is even true?

In response to criticism concerning the pandemic declaration, WHO Director-General Margaret Chan stated that “at no time, not for one second, did commercial interests enter my decision-making.”⁵ I believe her, and I also believe that she and her colleagues were following their own guidelines when they issued the pandemic declaration. However, what few of them may have appreciated is the extent to which those guidelines had been shaped over the past decade by the pharmaceutical companies that stood to profit from the declaration.