Junk science or Junk opinion?
December 27, 2006
There are differing opinions about whether future flu pandemics can be as severe as the 1918-19 pandemic. And as with religion or politics, some of those opinions come from shallow thinkers while other opinions come from more thorough thinkers.
Yesterday, I talked about the little cluster of reasons given for why "we will never see another 1918." The people who hold this view contend that "flu fearmongers" make misleading or false claims about the possibility of a severe pandemic. And these naysayers give reasons (motives) why fearmongers want you to believe a severe pandemic is possible. ("Flu fearmongers," you realize, would include the World Health Organization, the U.S. Centers for Disease Control and Prevention, national and local governments worldwide, leading virologists, veterinarians, and other experienced medical professionals, etc.) Yesterday, I discussed why I think the naysayers' explanations are poorly thought out.
(Bear in mind, the issue is not whether a pandemic will occur. Health experts are confident pandemics will continue to happen. The question is how severe future pandemics will be. Could they be like 1918?)
"Junk science expert" criticizes research in The Lancet
For a good example of how the naysayers can beat their drums, trying to debunk what they think are flu pandemic myths, look no further than a guest editorial at Fox News last week. Filed under the article category "Junk Science" is a piece entitled "One Flu Over the Cuckoo's Nest" by Steven Milloy. Mr. Milloy is described by Fox News as a "junk science expert." As an expert, he should know his stuff.
Milloy is highly critical of research just published in the prestigious British medical journal The Lancet. (The Lancet is one of the leading general medical journals, first published in 1823.) This new research attempts to estimate the mortality should a 1918-scale pandemic occur today.
The study was unlike earlier theoretical models which took a more qualitative approach, making strong assumptions about influenza attack rates and case-fatality rates. Instead, these researchers sought to conduct quantitative analyses based on actual government medical records from the time period. They only looked at countries which had nearly complete vital registration systems for the study period of 1915-23.
Their bottom line: Should a strain of influenza similar to what caused the 1918-20 pandemic emerge in our day, it could kill 51-81 million individuals (median estimate = 62 million).
The researchers were from Harvard University, the University of Queensland, and the Johns Hopkins Bloomberg School of Public Health.
Questioning researchers' motives and truthfulness
Milloy strongly questions the researchers' motives and methodology. Among the broadsides he levels against the researchers:
- He says "public health researchers can put more stock in frightening the public than informing it."
- He says the researchers' writings are "guesswork derived from other (cherry-picked) guesswork" and have "far more shock-value than credibility."
- He says they had an "apparent desire to come up with a scary – rather than a realistic – death toll from pandemic flu."
- He says their estimate of the effect of a 1918-like pandemic today is "the product of statistical modeling that uses worst-case death rate estimates from the 1918-1920 pandemic influenza."
- He closes with, "If we can't expect truthfulness and clear-thinking from public health researchers and officials in a time of relative calm, how much confidence can we have in them should a public health emergency develop?"
In sum, Milloy apparently thinks these researchers are bald-faced liars and data manipulators.
Same old claims; Same weak reasoning
Having rebuffed the research (and researchers) to his satisfaction, Milloy then cautions us that "the researchers ignored several key (not to mention glaring) differences between 1918 and 2006." Then he rolls out the standard naysayers' list of reasons why a severe pandemic simply can't happen in today's world. (If you saw my post yesterday, this will sound like a broken record.)
- "...modern medical care and public health practices have dramatically improved since 1918. So any flu epidemic is likely to be far less severe..."
- "...a great proportion of the deaths in 1918 was probably due to secondary bacterial infections that followed the initial viral infections. Today, antibiotics would be used to treat bacterial infections."
- "...much of the world was still recovering from the strains of World War I. Poverty, hunger, unsanitary living conditions and stress likely made much of the global population ripe for a killer flu pandemic."
- "...pandemic flu is quite rare and we ought to be cautious not to over-allocate scarce public resources to events that seldom happen and that seem to be getting less severe when they do occur."
- "...millions of people in the developing world die every year from preventable diseases, such as mosquito-borne malaria and diarrheal diseases caused by unsafe food and water. Why not spend available resources on saving people from those deadly diseases that are taking lives right now rather than on over-preparation for a hypothetical epidemic that is highly uncertain."
Naysayers' guide to "glaring differences" with 1918
OK, let's rummage through the above points and see if our "expert" has really exposed any "junk science."
Modern medicine will mute any pandemic.
Milloy says: "...modern medical care and public health practices have dramatically improved since 1918. So any flu epidemic is likely to be far less severe..."
As I explained yesterday, it doesn't matter if we have vaccine technology, oxygen supplies, ventilators, and other modern medical tools, if we can't deliver enough of these when and where needed. A pandemic causes a major surge in medical cases. The problem: our "surge capacity" is limited. We don't have enough oxygen supplies, ventilators, respirators, face masks - or doctors, nurses, and respiratory therapists.
Antibiotics to the rescue.
Milloy says: "...a great proportion of the deaths in 1918 was probably due to secondary bacterial infections that followed the initial viral infections. Today, antibiotics would be used to treat bacterial infections."
As I explained yesterday, especially virulent new flu viruses like the 1918 H1N1 are known to kill directly by primary viral pneumonia, not simply by secondary bacterial pneumonia. Antibiotics can't treat viral infections. Even for those cases where antibiotics could help, we again face the question of whether we'd have adequate supplies.
World War I made the world "ripe" for the flu.
Milloy says: "...much of the world was still recovering from the strains of World War I. Poverty, hunger, unsanitary living conditions and stress likely made much of the global population ripe for a killer flu pandemic."
Two points. First, a bulletin for Mr. Milloy: "Poverty, hunger, unsanitary living conditions and stress" are the daily norms in much of the developing world today. Remember? It's the same "developing world" which you later cite as being susceptible to malaria and other diseases (but not to influenza?).
Second, Milloy's description makes it sound like most of the world was a devastated battlefield after World War I. That's not the case. World War I was fought primarily in Central Europe, western Russia, and the Caucasus. There were limited activities in some Middle Eastern, African, and Southeast Asian areas, plus naval battles. Therefore, there were no war/post-war conditions in Asia, all of the Americas, the Scandinavian countries, most of Russia and the Middle East, and virtually all of Africa, Australia, and New Zealand.
Not that the war had no affect on any of these countries. (The United States and Canada, for example, sent troops to support the Allied Powers and suffered casualties.) But other than in specific geographic areas, the world was not a war zone whose people, infrastructures, and lands were ravaged by war. It's unjustified to say the war "likely made much of the global population ripe for a killer flu pandemic."
In contrast, a flu pandemic spreads worldwide. It doesn't understand borders.
Don't waste scarce resources on rare disasters;
Anyway, pandemics are getting milder.
Milloy says: "...pandemic flu is quite rare and we ought to be cautious not to over-allocate scarce public resources to events that seldom happen and that seem to be getting less severe when they do occur."
(1) I understand there needs to be balance and rational allocation of resources. But the "quite rare" disasters are some of the worst ones - the ones for which preparation is most valuable. Major earthquakes are "quite rare." Major tsunamis are "quite rare." Major hurricanes hitting major cities are "quite rare." But shouldn't we prepare for these?
(2) "Quite rare" means "low probability." We all know flu pandemics are low probability events. The problem is, they are also "high impact" events. Although they are rare, when they happen, they're awful. The potential for catastrophic losses makes planning vital. "The impact of this is so high that the risk rating tells you this must be a priority," says Don Ainslie, global security officer at Deloitte & Touche LLP.
(3) Although pandemics are low probability, they are certain to happen (unlike many low probability events). "Low probability" does not mean "never happens." It means the event is unlikely at any particular time. New Orleans may never again be hit by a major hurricane. But we will have other pandemics. Companies must understand that, "unlike many events that may never happen to a company, this is one that will," according to Michael Osterholm
(4) To say that pandemics are "getting less severe" is an incorrect conclusion. You cannot extrapolate the circumstances of the 1957-58 and 1968-69 pandemics. Those pandemics were sequentially less severe than the 1918 pandemic because the viruses involved happened to be less transmissible and/or less virulent. There's not a trend involved. It was just happenstance. Each pandemic is unique and depends upon the characteristics of the virus involved.
Spend on current diseases, not hypothetical diseases.
Milloy says: "...millions of people in the developing world die every year from preventable diseases, such as mosquito-borne malaria and diarrheal diseases caused by unsafe food and water. Why not spend available resources on saving people from those deadly diseases that are taking lives right now rather than on over-preparation for a hypothetical epidemic that is highly uncertain."
(1) Certainly, more effort and money should be directed at the infectious diseases mentioned. However, that does not mean we should ignore other infectious diseases like influenza which could kill perhaps tens of millions (or more) and disrupt economies worldwide.
(2) The very research article which Milloy cites suggests that 96% of the pandemic fatalities would occur in developing nations. Those nations don't need a flu pandemic on top of malaria and other diseases.
(3) I'm scratching my head over why Milloy calls a flu pandemic a "hypothetical epidemic that is highly uncertain." A nuclear war is hypothetical; it hopefully will never happen. Flu pandemics are not hypothetical. They happen. As Milloy just noted (!!), there were three pandemics in the last century. Why does he turn around and say any future pandemic is "hypothetical" and "highly uncertain?" The only questions are when the next one happens and how severe it will be.
Missing the point; more data = better findings
A fundamental problem: Milloy doesn't grasp why the researchers chose to study the 1918 pandemic instead of the 1957 or 1968 pandemics. He thinks they're up to something - "cherry-picking" the data and "manufacturing fantasy flu statistics" to support their fearmongering agenda. He writes:
First, the researchers' estimate of 62 million deaths has far more shock-value than credibility. It's guesswork derived from other (cherry-picked) guesswork.[snip]
But perhaps the researchers' choice that most reveals their apparent desire to come up with a scary – rather than a realistic – death toll from pandemic flu is their decision to use the 1918 pandemic flu data in the first place. There were, after all, two other more recent and, in all likelihood, more relevant pandemic flu outbreaks in the 20th century.
Milloy thinks the 1957-58 and 1968-70 flu pandemics would have been "more relevant" because:
"...it seems as though data from the Asian and Hong Kong flu pandemics would be much better starting points than the far more uncertain data from a chaotic period almost 90 years ago."
Well, that may be the way it seems to him ... but he's missed the point.
The express objective of this study was to provide a counter-balance to previous estimates - estimates which had been based on "strong assumptions about attack rates and case-fatality rates." The researchers sought to replace qualitative assumptions with quantitative information - "high-quality vital registration data," as they term it.
This criterion is even stated in the article's title: "Estimation of potential global pandemic influenza mortality on the basis of vital registry data from the 1918–20 pandemic: a quantitative analysis."
The researchers explain:
"...our aim is to assess vital registration data from the 1918-20 pandemic, since this pandemic was overwhelmingly larger than other 20th century pandemics and provides a clearly identifiable effect on mortality."
In other words, the 1918 event provides a much larger and clearer set of data (contrary to what Milloy thinks).
"Worst-case death rate?"
I've beat this drum before, but it's worth noting that Milloy's attitude is common among pandemic naysayers in this respect, also: they're running away from the concept of a really, really bad pandemic. They think 1918 was a worst-case aberration which should not be a key factor in evaluating the pandemic risk. Milloy says the Lancet researchers' estimate "uses worst-case death rate estimates from the 1918-1920 pandemic influenza."
All of us need to realize the 1918 pandemic wasn't necessarily a worst-case scenario. It was simply the worst of the more recent flu pandemics. That's probably why Milloy doesn't like it; as the worst of the last three pandemics, it yields a larger estimate for any modern pandemic. He seems to think the researchers picked the bigger number just to inflate their "scary" estimate.
Well, the researchers got even spookier. In this very article which Milloy is citing, the researchers expressly said they don't think the 1918-20 pandemic was necessarily the worst:
"In most discussions of influenza, the 1918-20 pandemic sets the upper limit, in terms of mortality, on what might occur in future pandemics. However, there is no logical or biological reason why that pandemic - albeit very severe - should represent the maximum possible mortality in a future pandemic."
Similarly, a leading insurance risk modeling company recently stated that the 1918 pandemic was not a worst-case scenario. In fact, they suggest there's a one in five chance we could see a pandemic more severe than 1918.
The flu virus of current concern - influenza A (H5N1) - almost exclusively causes bird disease. Should it develop a strain which causes human disease, we don't know how transmissible and virulent it would be. But the H5N1 bird virus has been described by perhaps the world's top flu researcher as the worst flu virus he's ever seen.
Taking a slap at "junk science"
Milloy says: "If we can't expect truthfulness and clear-thinking from public health researchers and officials in a time of relative calm, how much confidence can we have in them should a public health emergency develop?"
Milloy is taking pot shots at a report in The Lancet. If venerable journals like The Lancet and the New England Journal of Medicine aren't to be trusted, where can we go to find truthful, objective medical research on this topic? Where would Milloy suggest we turn?
More broadly, are the World Health Organization, the U.S. Centers for Disease Control and Prevention, the National Academy of Sciences, the European Centre for Disease Prevention and Control, the Pan American Health Organization, and many other world medical authorities mistaken to be concerned and warning people of the pandemic risk - indeed, urging individuals, communities, companies, and governments to prepare? Are all of the hospitals and other health care providers who are making pandemic preparations just mistaken? Do these experienced medical teams not know how to separate fact from fiction, with respect to the risk from flu pandemics?
What about the global manufacturers and service providers, the leading financial institutions, the multinational insurance companies, the top-tier business consultants - all of whom seek to maximize opportunities and minimize risks? They are evaluating the risks, closely monitoring developments, and making extensive preparations. Are all of these well-seasoned, highly-motivated people suffering from some collective delusion, when it comes to pandemic risk?
Are we to believe they're all being hoodwinked by "junk science?"
The naysayers talk as if those of us who fear and plan for a pandemic want a pandemic (and we, therefore, are disappointed or lose in some way if a severe pandemic doesn't materialize). Which is further evidence that the naysayers don't understand the matter - the unique scale and scope of a pandemic. If they understood the possible pandemic impacts that you and I foresee, they would know that a pandemic is nothing to wish for.
It is nothing to dismiss out of hand, either.
This post is part of a series: