Saturday, 11 October 2014

Ebola and the availability heuristic


A certain town is served by two hospitals. In the larger hospital about 45 babies are born each day, and in the smaller hospital about 15 babies are born each day. As you know, about 50 percent of all babies are boys. However, the exact percentage varies from day to day. Sometimes it may be higher than 50 percent, sometimes lower. For a period of 1 year, each hospital recorded the days on which more than 60 percent of the babies born were boys. Which hospital do you think recorded more such days?

The larger hospital (21)
The smaller hospital (21)
About the same (that is, within 5 percent of each other) (53)

The values in parentheses are the number of undergraduate students who chose each answer.

Most subjects judged the probability of obtaining more than 60 percent boys to be the same in the small and in the large hospital, presumably because these events are described by the same statistic and are therefore equally representative of the general population. In contrast, sampling theory entails that the expected number of days on which more than 60 percent of the babies are boys is much greater in the small hospital than in the large one, because a large sample is less likely to stray from 50 percent. This fundamental notion of statistics is evidently not part of people's repertoire of intuitions.

This example is drawn from “Judgment under Uncertainty: Heuristics and Biases. Biases in judgments reveal some heuristics of thinking under uncertainty.
Amos Tversky and Daniel Kahneman. Science, New Series, Vol. 185, No. 4157. (Sep. 27, 1974), pp. 1124-1131.

I am not particularly concerned about showing that our knowledge of Ebola is based on small sample sizes. That is a Good Thing. The problem we have now is that Ebola has achieved a very much larger sample size than in 1976 and subsequent outbreaks, as a result of, very probably, poorly controlled or non-existent public health services in a couple of dysfunctional nations, rather than any significant mutations in the virus, though mutations have certainly occurred in the past year.

I am trying to make a related point: commentators even a few months ago were being relaxed about the Ebola outbreak because the deaths at that time were, in total, far fewer than those caused by other diseases. The argument was that there were many other things to worry about. Malaria, for example. These reassuring voices proclaimed that this low transmissibility, high lethality disease would be brought under control, easily in the West, and with Western money and expertise eventually in West Africa. Avoiding panic was seen as a key step. “Car accidents more frequent than Ebola deaths, Ebola less transmissible than influenza” was the mantra.

As usual, the common folk were right to be alarmed. They noticed the death rate was high, and they didn’t like the sound of bleeding, vomiting and shitting themselves to death as their internal organs dissolved. Fussy, I know, but there it is. They also doubted that everything in Western health services would work perfectly, and that every member of staff would follow every protocol. Who does such things? as my elderly German car salesman expostulated, when I enquired whether, as proudly advertised, the rear seats of the car I wanted to buy could be removed to carry a large load.

Of course, the Kahneman and Tversky example could have been simulated, such that people eventually realised, as the large and small hospitals registered their births, that the results for the smaller hospital swayed around from day to day more than the larger hospital.

Below is a link to a simulation showing that in the early days Ebola is a slow burner, but then horrible deaths become more and more frequent and absolutely real, leaving influenza in the shade. Influenza is not the best comparator for Ebola.

Ebola spreads slower, kills more than other diseases

This simulation shows how quickly 10 diseases, from more fatal to less fatal, could spread from one person to 100 unvaccinated people.

So, just because an awful image is “available” to our attention, it does not mean that it is the sort of threat that educated folk should disdain.

By the way, Nigeria is doing something right. Gates had put in money to help them eradicate polio, so there were 100 trained epidemiologists on hand to use proper  techniques of infection control. Even so, they screwed up when their first case arrived (denying he had any contact with Ebola) and they lost 8 people before the staff rallied, overcame justifiable fear, and sorted it out. A highly creditable performance. Fingers crossed, because their health service is “fragile” and they cannot celebrate yet.


We will have to look at the numbers again every three weeks. Will Spain turn out to have lost more citizens than Nigeria?

Final, final point. The Washington Post has supplied a useful simulation. They also gave space to Laurie Garrett October 10 a senior fellow for global health at the Council on Foreign Relations. She attacks the view that a travel ban would keep Ebola from the USA. Link below, and below that her point No. 4 on travel bans.

The only evidence that any travel ban in the 21st century slowed down viral spread occurred right after the 9/11 attacks, when airports in the eastern United States were shut down for days, and few Americans travelled far from home for several weeks. Possibly as a result, the influenza season was delayed about two weeks in 2001. But the flu eventually came.

Many nations have banned flights from other countries in recent years in hopes of blocking the entry of viruses, including SARS and H1N1 “swine flu.” None of the bans were effective, and the viruses gained entry to populations regardless of what radical measures governments took to keep them out. The days of Ellis and Angel islands screening out diseases effectively disappeared with the jet age.

Influenza again. Garrett also assumes that the “jet age” is irreversible if lives are at stake.  She confirms that shutting down air travel probably delayed the influenza season by two weeks. Sounds a good deal to me. So, shutting down air traffic for a few days would delay Ebola by two weeks? Deaths from Ebola per 100 are 65 versus 1 for Influenza. The delaying air travel for several weeks, and from Ebola prone places until the outbreak is contained, should have an effect. In fact, I think she is saying that people value air travel more than disease prevention, and will continue to do so until the costs of disease become too high, at which time it will be too late to do anything. This is different from saying that “viruses gain entry to populations regardless of what radical measures governments took to keep them out”. We probably differ as to what constitutes radical, but it is clear that really radical measures would deny entry, or at least very severely reduce it.

It is up to the US to decide whether a visitor from Liberia must make his courtship visit to the US right now, or delay his ardour until West Africa has contained the outbreak.

Say it with flowers.


  1. As discussed by Greg Cochran on his blog, the total quarantine of American Samoa in 1919 resulted in zero flu cases or deaths, while the neighboring Western Samoa suffered the deaths of 20% of its population.

    1. Yes, acknowledged in a previous post.

  2. James, if you're so concerned about this disease, how do you account for the fact that Nigeria, a poor nation geographically connected to Ebola hot spots, has contained the virus? Ebola has been observed with R_0 values below 2 (see Chowell et al, 2004, "The basic reproductive number of Ebola and the effects of public health measures: the cases of Congo and Uganda" where they report 1.83 for Congo and 1.34 for Uganda.) I agree that strict measures should be enacted, because doing so is responsible and will give us good practice for when an actual epidemic appears on the horizon, not because I think there's any real possibility for my melting into an Ebola-riddled puddle.

  3. Sorry, but have you read this posting? I discuss Nigeria in this very piece. They are organised their sparse resources very well.

  4. Yes, I noticed Nigeria in this piece. Is that truly how you account for Nigeria's success? "There were 100 trained epidemiologists on hand?" How many trained epidemiologists are there in Spain? There are more than that number in Texas. Give this a few months; I really do think you'll see that Western nations, having advantages Nigeria can only dream about, will easily stop the Ebola virus.

  5. Dear Mark, Reading your blog, it seems we are in substantial agreement. Your link to Nigeria duplicates the main thrust of mine: Nigeria had epidemiologists and had practiced contact tracing. They took it seriously, and had success. I imagine, like you, that Western health systems ought to be able to contain the outbreak. However, they did not contain AIDS. Ebola should be much easier, but Spain did not make a good start. UK will be tested over the next few months. Also agree with you that if widespread jet travel reduces in number we will probably be far more effective at establishing quarantine. On the bright side, the last to be infected in any outbreak will be bloggers: they don't go out much. Best wishes.