Wednesday, October 1, 2014

The Mathematics of Ebola Trigger Stark Warnings: Act Now or Regret It

  • BY   

    Daliborlev (CC), FLickr
     Daliborlev (CC), Flickr
    The Ebola epidemic in Africa has continued to expand since I last wrote about it, and as of a week ago, has accounted for more than 4,200 cases and 2,200 deaths in five countries: Guinea, Liberia, Nigeria, Senegal and Sierra Leone. That is extraordinary: Since the virus was discovered, no Ebola outbreak’s toll has risen above several hundred cases. This now truly is a type of epidemic that the world has never seen before. In light of that, several articles were published recently that are very worth reading.
    The most arresting is a piece published last week in the journal Eurosurveillance, which is the peer-reviewed publication of the European Centre for Disease Prevention and Control (the EU’s Stockholm-based version of the US CDC). The piece is an attempt to assess mathematically how the epidemic is growing, by using case reports to determine the “reproductive number.” (Note for non-epidemiology geeks: The basic reproductive number — usually shorted to R0 or “R-nought” — expresses how many cases of disease are likely to be caused by any one infected person. An R0 of less than 1 means an outbreak will die out; an R0 of more than 1 means an outbreak can be expected to increase. If you saw the movie Contagion, this is what Kate Winslet stood up and wrote on a whiteboard early in the film.)
    The Eurosurveillance paper, by two researchers from the University of Tokyo and Arizona State University, attempts to derive what the reproductive rate has been in Guinea, Liberia and Sierra Leone. (Note for actual epidemiology geeks: The calculation is for the effective reproductive number, pegged to a point in time, hence actually Rt.) They come up with an R of at least 1, and in some cases 2; that is, at certain points, sick persons have caused disease in two others.
    You can see how that could quickly get out of hand, and in fact, that is what the researchers predict. Here is their stop-you-in-your-tracks assessment:
    In a worst-case hypothetical scenario, should the outbreak continue with recent trends, the case burden could gain an additional 77,181 to 277,124 cases by the end of 2014.
    That is a jaw-dropping number.