Age Structure and Case Fatality Rate

When claims are made that the new H1N1 virus is “mild”, we are often told that between 20,000 and 50,000 people die every year of  seasonal influenza. What we are rarely told is that 90% of these deaths occur in people over 65 years old (Foppa and Hossain, 2008). Further, people over 65 represent a relatively small percent of the total US population:

  • 0-14 years: 20.2%
  • 15-64 years: 67%
  • 65 years and over: 12.8%

Since most of the remaining deaths occur in babies and infants, this means that the case fatality rate for people in the 15-64 age range from seasonal flu is vanishingly small.

The new H1N1 virus is very different. The vast majority of the deaths have been in people under the age of 65.  Perhaps the most detailed calculation of case fatality rates by age was provided in a leaked CDC Director’s brief (from July 16, 2009) that was apparently uploaded to Cryptome. According to this document, here are the case fatality rates for different ages (n = 262):

  • 0 -4 years: 0.17%
  • 5-24 years: 0.22%
  • 25-49 years: 1.5%
  • 50-64 years: 3.33%
  • 65 years and over: 5.24%

In the Spring and early Summer, most of the cases in the US were in children because the virus was spreading rapidly in schools and summer camps. Note, the case fatality rate for this age range is 0.22%. However, in the Fall, we can expect that although the virus will likely primarily affect school children in the beginning, it will eventually spread to the older age groups.

Based on the leaked information on case fatality rates from the CDC and the known population structure of the US, it is reasonable to project that the unmitigated case fatality rate for the US in the Fall/Winter will be much higher than in the Spring, even with no change in the killing power of the virus.

For countries that lack Tamiflu or vaccines, a 1918 level pandemic is entirely possible.


4 thoughts on “Age Structure and Case Fatality Rate

  1. We don’t have confidence in the denominator, do we? I think it was you who once pointed out that CFR, actually severity in general, is not predictive. We don’t know the severity while it is going on.

    With the CDC’s no testing policies it is as if we are back in the days before testing was available. Just as with 1957 we won’t know the severity numbers – attack rate, hospitalizations, deaths – until they’re released at a conference a year or two from now.

    In theory couldn’t we do seroprevalence tests on a random sample of people and extrapolate the results to the general population? Once we have an idea how many people have had an H1N1 infection, we’ll have a denominator we can have confidence in. If the HMO reporting organizations checked any blood draw for H1N1 antibodies the last week of AUgust, and blinded the results, wouldn’t that give us an idea of prevalence in the general population?

  2. The CDC did seroprevalence studies and never released them, right?
    Guess it wasn’t good news.

    The US lacks Tamiflu
    (the “goal” was to never have any for 75% of the US population;
    it wasn’t even enough for one wave –
    some states didn’t even bother buying 25% coverage – despite being in Alert for H5N1, which was and is fatal without it,
    The US has been donating Tamiflu before and during pandemic to other nations, yet Americans are dying for lack of honest diagnosis & antivirals at symptom onset)
    and, the US lacks effective vaccines, in quantity, in time.

    The WHO and the CDC knew the CFR’s by age decades from Mexico early on (see results in the WHO weekly report May 22);
    they could have called far worse than a, “Cat. 2” Pandemic – and then, the US didn’t even use the Community Mitigation Non-Pharmaceutical Interventions, “planned” (our, “only tool” without vaccine and enough antivirals) for “Cat. 2”!

    39% of the confirmed reported US deaths
    (“died at home of natural causes” adults are Not getting tested nor counted)
    as of 16 July, in that CDC Director’s Brief,
    were ages 25 to 49.

    39% of deaths are grad student, parents of infants, students’ parents, reproductive age women
    (pregnant infected have an 8 times greater risk of death than average) and, those ages are also our Essential & Critical Infrastructure workers.

    39% of the dead in the Spring/Summer were ages 25 to 49;
    if this, “expected” ramp-up in cases, now that summer daylight is fading and crowded schools are forced to stay open,
    causes 3 to 5 times as many cases as places had at the beginning of Pandemic…

    obviously the medical system (and communities) will be overwhelmed, and the fatality rates will all rise (including from non-panflu causes).

  3. billp, actually, we do have the denominator for the case fatality rate. It is the number of clinically significant “cases”, not the total number infected. The estimates for CFR for seasonal flu and previous pandemics were made on the same basis. So, although there may be individuals who are infected but don’t exhibit symptoms, they weren’t counted in any of the pandemics nor for seasonal flu. Thus, the CFRs provided above are completely comparable with seasonal or previous pandemic CFRs.

    If one wanted to make the case that there are very large numbers of “mild” cases that have been missed, seroprevalence studies would be the way to test that hypothesis. The CDC has had the means to do this for months. They have also had a “team” down in Mexico that was supposed to be doing this, months ago. Yet, AFAIK, the results were never released.

    There really is no reason to believe that we are missing a large number of mild cases as compared to seasonal or previous pandemics. The numbers are what they are.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s