A different Ebola plot shows that number of cases is stable since one month

14 Oct

Note 2 (also Oct 14): Christophe Fraser (@ChristoPhraser) pointed out that a short article in the Lancet in 2003 about SARS made exactly the same point as I do here. 

Note (added Oct 14): several people on Twitter told me that just now, the fraction of cases that is reported is going down, meaning that the data become more and more unreliable. I don’t know if this is true, but if it is, then no matter how we plot the data, they may not mean much.

The other day, when I was writing about Ebola and airport checks, I noticed that one of the graphs on Wikipedia showed the number of new cases per week in stead of the total number of reported cases. Suddenly, the epidemic looked very different to me! Even though new cases are being found each week, the Wikipedia graph showed that the number of cases per week had been stable for a while. This is good news! Of course, we need the number of new cases per week to go down (and maybe it has just done that … says at least a Dutch newspaper), but even if the number is no longer going up, that is already a great improvement. I decided to get some data from Healthmap using code by Ebolastats from their Github repository, and plot the number of new cases per day, averaged over roughly weekly periods.   To me it looks like the number of new cases per day has been fairly stable since the week ending with September 9th.

CasesPerWeek

At the same time, if we use the same data to plot the total number of cases, it looks like the number is still increasing exponentially. However, if you just focus on the last 5 points, you may agree with me that you could fit a straight line through it.

Note that the total number of cases cannot go down. When the epidemic is over, this curve will flatten to a horizontal line, it will not go back to the x-axis. I think most of us are used to looking at plots that show the number of cases per week or per month and so, even without realizing it, when we see a curve that is going up and up, we think that it means that the epidemic is getting worse, even if it is really stable. Therefore I don’t like the kind of plot that shows the total number of cases, because I think it looks much worse than it is.

TotalCases

Advertisements

2 Responses to “A different Ebola plot shows that number of cases is stable since one month”

  1. Erik October 15, 2014 at 6:23 am #

    sorry, don’t agree, teams on site are just at their maximum capacity to record. that’s why the numbers of cases/week don’t go up anymore. but indeed people expect the graph of total case to go down as the epidemic is stopped wich it will never do ofcourse. this is wat WHO says on october 8:

    The past week has seen a continuation of recent trends: the situation in Guinea, Liberia, and Sierra Leone continues to deteriorate, with widespread and persistent transmission of EVD. Problems with data gathering in Liberia continue. It should be emphasized that the reported fall in the number of new cases in Liberia over the past three weeks is unlikely to be genuine. Rather, it reflects a deterioration in the ability of overwhelmed responders to record accurate epidemiological data.

    http://apps.who.int/iris/bitstream/10665/136020/1/roadmapsitrep_8Oct2014_eng.pdf?ua=1

Trackbacks/Pingbacks

  1. Newer Ebola plot shows number of reported cases per day is stable | Being A Better Scientist - October 24, 2014

    […] I just downloaded the latest Ebola numbers (data from Healthmap using code by Ebolastats from their Github repository) and the number of reported new cases per day is still stable (see my earlier post on why plotting cumulative cases makes the situation look worse than it is) […]

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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

%d bloggers like this: