Skip to content

why I think the fatality rate in the current Ebola epidemic suffers from undercounting

30 July 2014

Two days ago the World Health Organization (WHO) released its most recent update on the Ebola situation in West Africa. Here are the numbers.


One debate about Ebola numbers has been about journalists’ use of a particular statistic: the fatality rate from Ebola. In the worst epidemic ever, 90% of those reported infected died from Ebola (and this is the rate you’ll see cited in many of the news stories about Ebola). The chart of fatality rates since 1976 shows a range from 25-90% for any epidemic with more than one reported Ebola case. This helpful image of Ebola fatality rates created by Brett Keller excludes epidemics with fewer than 30 deaths:

Analysts debate whether using the 90% fatality rate is an effort to dramatize the disease outbreak. According to the WHO’s recent report, the fatality rate is at roughly 56% (divide 672/1201) for reported cases in the current epidemic. So some analysts are reporting this lower figure, citing it as more accurate. But Carl Bialik at FiveThirtyEight points out that the fatality rate in the current Ebola epidemic is just an estimate:

…there are several reasons to interpret the reported death toll as very preliminary; some reasons mean it might rise from here, while others mean it might fall.

He offers four reasons why the fatality rate is a moving target. In brief:

  1. The outbreak isn’t over — there are still many people who are sick and likely to die. Even with no new cases, as the disease runs its course in those already infected, the fatality rate will go up.
  2. False positives (meaning suspected or probable not yet confirmed cases of Ebola that will eventually be ruled as not Ebola) temporarily inflate the number of people infected, making the overall ratio of deaths to infected seem lower than it really is.
  3. Some people may never get tested for Ebola because they recover. Some Ebola symptoms are also symptoms of other diseases (e.g., malaria, flu, etc.) and so these people could self-diagnose as having something else. Those with more severe symptoms might be more likely to seek treatment.
  4. People who die from Ebola without diagnosis will eventually be counted in WHO statistics.

First, I haven’t heard of people suffering from Ebola that mend on their own. (Anyone have more info on this?)

But more importantly, I think it’s most likely that there will be an undercount of deaths and that the fatality rate will be an underreport. Bialik is confident from his conversation with a WHO spokesperson that eventually people who died from Ebola without seeking treatment will be counted. I am not so confident.

With some communities fearful and suspicious of clinics treating Ebola, it is no doubt there are Ebola cases going unreported. As these people die from Ebola (the fate for many of them), what incentive do their families have of alerting health authorities after-the-fact? How could health authorities posthumously test a suspected Ebola case — especially following burial?

5 Comments leave one →
  1. 30 July 2014 1:42 pm

    I don’t know if the WHO has released enough data to calculate, but you can take care of most of problem #1 by excluding the last few weeks of new cases, but keeping all the deaths. Then you’ll have a reasonably accurate fatality rate.

    • 30 July 2014 1:47 pm

      That would get us closer.

      The real denominator should be the sum of deaths and those who have recovered.


  1. Ebola patients are real people, deserving of more | haba na haba
  2. Ebola Update: 729 reported dead; Sierra Leone declares emergency; WHO leverages $100m response - Humanosphere
  3. Ebola is only the Kardashian of diseases if you think Africa is a country | haba na haba

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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s

%d bloggers like this: