Avian flu sounds pretty nasty, and a pandemic would be a disaster. But John Sutherland, “writing in the Guardian”:http://www.guardian.co.uk/g2/story/0,3604,1462141,00.html , is in the grip of statistical confusion when he asserts that it could kill 70 per cent of the population. As I understand it, the virus kills 7 out of 10 people that it infects, and the number infected is far below 100 per cent. Moreover, the 7 out of 10 figure may well be an exaggeration, since people who recover and don’t die are less likely to be be included in the figures than those who do. The WHO “impact assessment”:http://www.who.int/csr/disease/influenza/preparedness2004_12_08/en/ isn’t encouraging (2 to 50 million dead, but could it be worse than that). I’m sure we have some epidemiologists among our readers. Any thoughts?
{ 1 trackback }
{ 9 comments }
dsquared 04.18.05 at 1:32 am
Entirely correct Chris. Note also that there is some interplay between lethality and infection; diseases which kill everyone who gets them very quickly, don’t tend to spread all that fast, even if they kill them in such a way as to make them outlandishly infectious (like Ebola). The projection for the UK is, IIRC, 50K at the low end and c2m at the high end.
(also note that the 70% figure has been put together based mainly on the effects of the disease in Vietnam; people in the UK are in general better nourished and in better overall health)
Neil 04.18.05 at 1:39 am
The virus (because of the lack of error correction machinery in the way it replicates) evolves very quickly. So if it infects lots of people, we’ll see competition between various strains of it. We can expect less lethal (or at least less quickly lethal) forms of it to win out over those that kill their hosts quickly (since hosts who are quickly incapacitated don’t spread the virus as effectly as those who keep going). So I would expect that 70% figure to drop fairly rapidly.
derrida derider 04.18.05 at 2:49 am
The mechanism neil points to is the way all pathogens tend to go – for example, over time we can expect HIV to be less virulent (in fact, its already happening). The most successful pathogens spread quickly and kill slowly.
But that’s small consolation to those who get the disease while it’s still virulent.
Michael Otsuka 04.18.05 at 3:20 am
As a general rule, I don’t read what newspaper columnists write unless I have reason to think they know what they’re talking about. So I’ll read Krugman on economics but not on politics, and a Professor of English writing on epidemology would be be pretty near the bottom of my list.
But, in Sutherland’s defence, his academic web page says that he’s “currently engaged on the Oxford Companion to Popular Fiction”, so maybe we should treat this as a work of science fiction.
mw 04.18.05 at 6:33 am
The most successful pathogens spread quickly and kill slowly.
Actually, the most successful pathogens kill not at all. In fact, the most successful pathogens incapacitate their ‘hosts’ as little as possible (an incapacitated host does a lousy job of wandering around and spreading the virus). Pathogens that kill most of those infected invariably are ones just crossed over from another species (in which species they are far less virulent), and once they become established in humans they tend to ‘learn’ how not to kill them fairly quickly.
(One exception to this, BTW, is anthrax which can only spread by killing its host).
Steve LaBonne 04.18.05 at 7:44 am
Most newspapers apparently require prospective reporters and columnists to prove that they are innumerate and scientifically illiterate before they can be hired.
eudoxis 04.18.05 at 10:20 am
It’s not at all certain that the avian flu kills 70% of its hosts. The epidemiological studies of this flu have been fraught with mismanagement, incomplete testing, misdiagnoses, etc. In any event, it’s true that the host mortality rate doesn’t apply to the population at large.
There are some misconceptions above in the comments about the relationship between host mortality and success of the virus. The most successful viruses are those with high transmission rates. In general, high transmission rates place a greater burden on the host by using up its resources leading to greater mortality. For air-transmitted viruses, higher mortality rates are positively associated with higher transmission rates. There are no hard and fast rules, of course, and the evolutionary relationship between the host and the virus is complex. Eventually, a population may evolve to become resistant to the virus.
The very long term success of a virus is really irrelevant to the evolution rate of a present virulent strain. There are instances when benign strains are selected for to enhance transmission rates in cases where intermediate vectors and extra time are necessary. Short term reproduction is associated with success and it doesn’t matter if a virus would eventually destroy its host population and grow extinct with it. There is no particular tendency for the virus as a group to think about their evolutionary long term success (contra Gould).
Steve LaBonne 04.18.05 at 10:25 am
Much more on such questions can be found in Paul Ewald’s classic book “Evolution of Infectious Disease”.
lemuel pitkin 04.18.05 at 10:34 am
Second labonne’s recommendation of Ewald. One point he makes is that it is not treu taht pathogens in general evolve toward reduced virulence. The key question is transmission. Pathogens transmitted directly from host to host evolve toward reeduced virulence beceuase mobile hosts are far more effective vectors than immobile ones. Pathogens transmitted by insect vectors or water are a different story. An incapacitated host may well be mroe likely to be bitten by mosquitoes, etc. And if you’re a water-borne pathogen, chronic diarrhea is the way to go.
One interesting argument he makes is that the 1918 flue epidemic was the result of conditions on the Western Front, where crowding and field hospitals allowed incapacitating flue virus to spread successfully, rewarding the most virulent strains.
Comments on this entry are closed.