Effect Measure

Mild pandemic? Bite your tongue.

A spot-on column in CIDRAP Business Source [subscription] by Center for Infectious Disease Research and Policy’s (CIDRAP) Director, Mike Osterholm, reminded me to say something I’ve wanted to say for a long time. We should banish the word “mild” from the influenza lexicon. There’s no such thing as a mild case of influenza, any more than there are “mild” auto accidents. There are cases that for reasons we don’t understand don’t make you very sick (or sick at all), and there are cases that can lay you lower than you ever want to be, including six feet under. What Osterholm does with great cogency is put paid to the idea this is a mild pandemic. His reasons will be familiar to readers here, but he says them extremely well.

Both Osterholm and we find CDC’s most recent estimates (November 14) of about 10,000 deaths, 47 million infections and over 200,000 hospitalizations “credible and thoughtful.” To compare the oft quoted number of 36,000 excess deaths from seasonal flu to this 10,000 number is an “apples and oranges” affair. The 36,000 number is an excess mortality figure derived by different and non-comparable methods (see our post here for more details). Here is some of Osterholm’s version:

In that CDC study, only 9,000 of those estimated annual seasonal deaths are due directly to influenza or secondary bacterial pneumonia. The other deaths are among persons who have influenza and who die of events like heart attacks or strokes. If you want a comparison, think of the guy who has a heart attack while snow blowing his driveway after a large snowstorm and whose death is labeled “storm-related.”

More important, though, is what we pointed out early on. It’s not just the number of deaths but the pattern of illness in the population, flu’s descriptive epidemiology:

More than 90% of the estimated seasonal influenza deaths occur in the elderly, who in many instances have existing serious health conditions that mean their deaths may not be far off, regardless of their influenza illness. We all realize that death is inevitable, and, as a public health practitioner, I find that this mad race to eliminate the top 10 causes of death is not always well thought through. If we were to accomplish such a goal, there would be 10 new leading causes of death, and I’m not so sure some of those would be better than the current ones. But I think we can all agree that “early deaths”?or those that occur well before our elderly years ?just shouldn’t happen. The way we count influenza mortality, an influenza-related death in an 87-year-old person with advanced Alzheimer disease is the same as the death of a 22-year-old otherwise perfectly healthy pregnant woman. Both deaths are equally tragic, but any reasonable person would agree they are not equivalent public health outcomes.

[snip]

Of the estimated 9,820 deaths:

  • 1,090 (11%) have occurred in children 0-17 years of age
  • 7,450 (76%) in people 18-64 years of age
  • 1,280 (13%) in people over 65 years of age

This age distribution differs considerably from what we see with seasonal influenza. (Dr. Michael Osterholm, Featured article CIDRAP BUsiness Source; subscription required)

Which brings me back to the issue of “mild” flu and mild pandemics. While this pandemic is not 1918 (for which we can be grateful, although no one knows what prevents this or any other pandemic from repeating that catastrophe), it is still pretty bad, already producing more estimated hospitalizations than seasonal flu and we aren’t even into the heart of flu season yet:

It’s challenging our healthcare system unlike any previous seasonal influenza season over the past 30 years. That makes it hard for me to call this a “mild? pandemic. I just don’t believe the term works.

Yes, there are mild, moderate, or severe influenza illnesses on an individual basis. But how do we describe a pandemic that hits a limited group of people really hard and causes only “routine influenza” for most others?

Whatever you describe it, I’m with Osterholm. Don’t call anything connected with flu, mild. Most flu infections don’t have dire consequences, but a significant number do. And we never know who is going to win the lottery. The convergence in our views extends to the same analogy: automobile accidents. As we’ve noted several times, it doesn’t make much sense to call any encounter with several tons of moving steel “mild.” It may be an encounter that produces little or no damage, but that’s luck. Anyone who’s gotten the repair bill for what can happen when hit by another car going only 5 miles per hour won’t consider the encounter “mild.” More importantly, any such encounter has the potential to be a catastrophe and the really bad ones are often just plain dumb luck. Mostly we don’t have such terrible luck. Yet even a fender bender can be expensive or result in minor injuries and be emotionally traumatic as well as result in lost work. Whatever you might be tempted to call an auto accident (minor or a scratch or a fender bender), very few people would be prone to call it “mild.” And whle the number of people killed each year in motor vehicle accidents is roughly the same as seasonal flu, we don’t usually think of our nation’s annual highway death toll of 40,000 as “mild” either. Its age distribution isn’t too different than pandemic flu’s either, the source of much anxiety to any parent whose child has just gotten their driver’s license.

It’s not just that “mild” is an inapt word. It is an inapt and dangerous idea. It is based on comparing total deaths in flu seasons with each another, not with the public health toll they exact. It is worse than a word not conveying the proper seriousness. It sends a message that itself has consequences, promoting a lack of urgency about taking rational public health measures like vaccination.

That’s especially pertinent this year when we don’t yet know what the usual flu season (January to March) will bring, either with seasonal flu or a recrudescence (“wave 3″) of swine flu. As Osterholm points out, even with available vaccine now certain in the next few months, we are likely to find ourselves in January with a very large proportion of the population still without immune protection.

After all, what’s the hurry? It’s a mild pandemic.

Comments

  1. #1 Greg Laden
    December 16, 2009

    The 36,000 number is an excess mortality figure derived by different and non-comparable methods

    Is there, or will there be, an estimate of 2009 H1N1 (nee Swine) flu death rates using methods such as Serflings/Thompson’s in order to make this comparison?

  2. #2 revere
    December 16, 2009

    Greg: Yes. The methods continue to be refined but there has to be sufficient data to fit the seasonal curve to calculate the residuals (to get the “excess” associated with flu season) so that can’t be done as we go along (I believe it’s because there are too few points at the end of the data stream, just as for usual regression). So we’ll have to wait for that.

  3. #3 SusanC
    December 16, 2009

    While I agree with the gist of your assessment, especially the ‘it ain’t over till it’s over’ part, I have some caveats. First, there’s an inherent comparative that underlies the word ‘mild’. Always the question – spoken or not – is, as compared to what?

    My son is a cyclist – and I don’t mean the “can I ride around the park, mom?” kind. If he thought he had any chance, he would have gone for the Tour de France. The roads in England are crowded, narrow, and these days filled with over-sized vehicles driven by folks who (I’d like to think) are mostly mild, polite, and considerate – until they get behind the wheel. So when he got hit by a van, (at slow speed at a junction) and suffered nothing more than soft tissue injury to his legs, I was totally relieved. Yes, he was on crutches for weeks, but at least there were no broken bones and no injuries to more vital parts such as the knees or pelvis or worse. As far as being-hit-by-a-van-while-you-are-on-a-bike goes, I’d call it a mild/minor accident, probably a good cautionary experience that made him a lot more road savvy since.

    We are also a family of skiiers. Like many skiiers my age, I have this rather jaundiced view that all snowboarders are suicidal/homicidal maniacs. I know the vast majority enjoy their sport with no harm, but I have 2 friends with kids who suffered snowboarding accidents – one a 9-year old who fractured his wrist one year, and another whose 20-something son is now a quadriplegic. I myself broke my shoulder (specifically, the upper arm bone, the humerus) one year – recovery was painful because it wasn’t the kind of fracture you can set and put in a cast. Rather, you go through 3 months of wearing a sling such that any little jostle causes the bone ends to grate upon each other, or so it seemed. But was it ever dangerous, life-threatening, or something to worry about? No, it wasn’t.

    Are there different grades of skiing/snowboarding/road accidents? I’d say, very much so, all the way from minor scrapes to quadriplegia, even within my small circle of friends/family.

    The point I’m trying to make is, there ARE different grades of severity for pandemic flu. The problem that I see, especially now looming upon us, is that for the vast majority of the public, they DO experience this as mild. The risk in portraying it differently lies in losing your audience, with the accompanying damage to your/our credibility. Already, justified or not, there are accusations of scaremongering. My concern is that unless we can portray the risks in ways that are meaningful to the public, we’ll lose any momentum that we ever had, for pandemic preparedness for something much more devastating than the current pandemic, such as the still-current threat of H5N1.

    I believe the solution lies in stepping away from the simplistic and dichotomous description of severe vs mild, into a more sophisticated message that incorporates the tremendous variations possible with any pandemic (eg about 100x difference between 1918 and now, and likely much much worse if H5N1 were to cause the next one…), our inability to predict what the flu virus is capable of, and just because we didn’t have ‘the big one’ this time round, doesn’t mean that the next one won’t be, etc..

    I believe given the chance and the right message, the public ARE capable of learning from their experiences. Just as my son learned that roads are potentially dangerous places to ride on and has become a lot more ‘defensive’ in his road-tactic (aka more risk-aware and ‘prepared’ in the pandemic lexicon), I believe the message should be,

    a) it ain’t over yet,
    b) we are reasonably lucky this time,
    c) but we are unlikely to be just as lucky forever, and
    d) now is the time to reflect on what else could have happened (or may still happen) if things were different, so that we can make sure to put in enough resources into the system, for next time.

  4. #4 Rob Jase
    December 16, 2009

    You are absolutely right SusanC, its way too early to judge the results of H1N1. It hasn’t acted like a regular seasonal flu so far & I wouldn’t bet against it showing some differences in the longer term.

  5. #5 biologist
    December 16, 2009

    Revere & Osterholm:

    Exactly. Thank you.

    In addition to the fact that flu kills every year and is a nasty disease that shouldn’t be labeled “mild” in general, it’s also time for us to stop using adjectives referring only to case fatality ratios to describe entire pandemics. We need new terminology that takes into account age distributions for morbidity & mortality as well.

  6. #6 Sid Offit
    December 16, 2009

    According to the CDC:

    The estimated ranges generated by this methodology provide a sense of scale in terms of the burden of disease caused by 2009 H1N1. It may never be possible to validate the accuracy of these figures. The true number of cases, hospitalizations and deaths may lie within the ranges provided or it’s also possible that they may lie outside the ranges.

    —————————
    I think its a mistake to throw around the 10,000 figure as if it reflects reality. Especially when the organization estimating deaths is the same one behind the pandemic hype.

  7. #7 Patricia Pitsel
    December 16, 2009

    Language is a shifty thing – and words, Lewis Carrol notwithstanding, can mean whatever we want them to mean. When we say a pandemic is mild, do we mean that there are fewer fatalities than anticipated? A lower CFR than in previous pandemics? Fewer people infected? People recover faster? Less economic disruption?

    Not only do we need to be clear what our terms mean, and as stated earlier, “mild compared to what”, we do need to educate people to understand risk.

    My experience teaching in Canadian universities and colleges suggests that a number of students are not even sure of how to calculate percentage unless they are majoring in maths or science. It’s akin to the old joke about a couple with 4 children who decided to stop since they had heard that every 5th person in the world was Chinese.

  8. #9 Sam Dawes
    December 16, 2009

    I urge everyone to look at the PBS special “Anatomy of a Pandemic,” which is airing this week and is on the http://video.pbs.org website. It makes it quite clear that this year’s H1N1 epidemic is anything but mild in terms of its global impact. It also has side interviews, including one with the Director of the Boston Public Health Commission, Dr. Barbara Ferrer. She points out the differential impact of H1N1 on citizens of Boston who are African-American or Latino. By almost any lens, this illness is taking a large toll.

  9. #10 anon
    December 16, 2009

    You cannot change English language and people will continue calling it mild.

    it’s not only not 1918, it’s also not 1957 nor 1968.
    This is the mildest pandemic than we know of had ever occurred.
    It may be not so mild when compared with seasonal flu,
    but it is mild when compared with previous pandemics or when
    compared with earlier estimates (Ferguson compared it with 1957,
    PCAST gave 30000-90000 deaths)

    Wasn’t it worth it, if it now replaces seasonal H3N2 ?

    the 10000 estimate inludes those deaths from 2ndary cases,
    I assume it’s based on the 122-cities mortality data.
    there will probably be much fewer total seasonal
    excess-deaths this season than normal.

  10. #11 TyMbuktu
    December 16, 2009

    Re anon post above. Don’t you mean. ” yet”?

  11. #12 pft
    December 16, 2009

    From CDC web site.

    “Most people with 2009 H1N1 have had MILD illness and have not needed medical care or antiviral drugs, and the same is true of seasonal flu.”

    Also this:

    “Currently available rapid influenza diagnostic tests cannot distinguish between 2009 H1N1 and seasonal influenza A viruses. Most people with flu symptoms this season will not require testing for 2009 H1N1 because the test results usually do not change how you are treated.”

    Fine, testing not required. But…

    ” As of September 2009, more than 99% of circulating influenza viruses in the United States are 2009 H1N1.”

    How do you know this if you don’t test?

    “CDC has recommended that this testing be focused on 1) people who are hospitalized with suspected flu; 2) people such as pregnant women or people with weakened immune systems, for whom a diagnosis of flu will help their doctor make decisions about their care.”

    Ok, so they only test those who are really sick or have underlying conditions. Hopefully they do not extrapolate the outcomes from this sample to the population at large to “estimate” deaths from H1N1.

    Also, rather than an 18-65 grouping it would be nice to see a more detailed breakdown since it may be a larger number in the 50-65 who have health problems (obesity, diabetes, etc)

    But having heard stories of the 1918 pandemic in Boston from grandparents, I am pretty sure they would categorize this flu as mild. And as my grandfather used to say, “what doesn’t kill you makes you stronger”.

  12. #13 revere
    December 16, 2009

    pft: They have a virological surveillance system. Go to the CDC site and you can read about it via FluView. And the age groupings are finer there.

  13. #14 Clark@amanochocolate.com
    December 16, 2009

    Pft there are two tests – one that is quick and inaccurate and an other one that takes a day typically (assuming no backlogs – my test took 4 days to get back to me)

    While I agree with some of the pushback that there was too much fearmongering, h1n1 has been serious. I was hospitalized (much to my shock) for two weeks with it! (After suffering a week at home) I’m still on oxygen. I was shocked that there was a full floor with people with h1n1 not including all those in intensive care. (Where I almost ended up) I’m home but am still on oxygen and still unable to work.

    This is not your typical flu!

  14. #15 Ron Law
    December 16, 2009

    For all intense and purposes It’s over… Take a look…
    http://www.cdc.gov/flu/weekly/weeklyarchives2009-2010/images/image481.gif

    http://www.cdc.gov/flu/weekly/weeklyarchives2009-2010/images/EIP48_small.gif

    You quote, “Of the estimated 9,820 deaths:

    * 1,090 (11%) have occurred in children 0-17 years of age
    * 7,450 (76%) in people 18-64 years of age
    * 1,280 (13%) in people over 65 years of age”

    As far as I can tell these are not valid numbers… they have taken syndromic pneumonia deaths (total) and redistributed them based on the proportions of deaths for confirmed H1N1… http://www.cdc.gov/h1n1flu/tools/impact2009/H1N1_Impact2009_v1.xls

    That is not valid… CDC must publish the age distribution of the syndromic pneumonia deaths… most claimed flu deaths are due to this… and nearly all occur in elderly people… If syndromic pneumonia was killing thousands of children then the hospitals & esp ICU’s would be cluttered with such patients. I have seen zero reports that that is the case. Until CDC publish age breakdowns of the syndromic pneumonia deaths their results have the look of a ‘black box’ origin.

    Ron Law

  15. #16 Julie Stahlhut
    December 16, 2009

    Gaaahhhhh. Even “mild” flu makes you feel like hammered crap for a few days. I don’t understand people who won’t get immunized simply because most H1N1 victims get better in a few days. It seems like not bothering to properly refrigerate your leftovers because most cases of food poisoning only make you throw up and have explosive diarrhea for a few hours.

  16. #17 Paula
    December 16, 2009

    Clark, good to hear you are home and avoided ICU. Be well soon.

  17. #18 jerah
    December 17, 2009

    My husband and son just had the flu a little over a week ago. We’re assuming it was H1N1, just because there doesn’t seem to be much else around. They seem to have recovered well (I’ve given them plenty of recovery time, as well as a bunch of other herbal remedies, vitamin , etc), but our friends, who have 3 young children (under 3) were not as lucky.

    They had the flu about a month ago and, after a series of “colds”, ear infections and gastrointestinal distresses, the father was hospitalized for 2 days last week with bacterial pneumonia. Their youngest (a 1-year-old) was hospitalized today with pneumonia. There’s nothing mild about this flu for them.

  18. #19 elephantman
    December 17, 2009

    anyone who thinks we are experiencing pandemic that can be accurately characterized as “mild” need only look at the pediatric mortality graphs published in CDC’s Flu Weekly update to disabuse themselves of this fallacious notion.

    http://www.cdc.gov/flu/weekly/weeklyarchives2009-2010/images/IPD48_small.gif

  19. #20 anon
    December 21, 2009

    I don’t understand you people.
    Of course it’s still bad and not mild when compared
    to food-poisening or common cold.

    But it is mild when compared with other pandemics.
    Or when compared with what was expected.

Current ye@r *