Thursday, 14 November 2013

The book of MERS has several chapters yet to write

Epidemic is a big word, and while it generally means "a rise in the number of cases above what you'd expect", you can see from the definitions below that there are many ways to spin the meaning. For the public at large, it generally means "bad scary stuff" and so it's important that we use this word sparingly.

An epidemic is defined by Oxford Dictionaries as:


a widespread occurrence of an infectious disease in a community at a particular time

..or more applicably..



a sudden, widespread occurrence of an undesirable phenomenon

...from Merriam Webster online...


affecting or tending to affect a disproportionately large number of individuals within a population, community, or region at the same time

...from Wikipedia...


In epidemiology, an epidemic (from επί (epi), meaning "upon or above" and δήμος (demos), meaning "people") occurs when new cases of a certain disease, in a given human population, and during a given period, substantially exceed what is expected based on recent experience.

The Middle East respiratory syndrome (MERS) was so-named back in May 2013, and prior to March 2012, there had been no known cases of the coronavirus (CoV) named for the disease it was associated with.

Yesterday we saw a detailed publication by Cauchemez and colleagues in the Lancet Infectious Diseases (LID). Accompanying that was an excellent piece in the Canadian press written by Helen Branswell which included some comments from the authors.

Click to enlarge. 
Accumulation of MERS-CoV lab detections by week (blue
mountain, 
left y-axis) and the accumulating deaths
(red line, left y-axis). The proportion of fatal cases is slowly
declining as fewer cases have died recently (ratio; black line,
right y-axis). No data exist for ~3 or so deaths and I include
the unconfirmed 2nd case in Kuwait for now.

Feel free to use, just cite me and here.

The key phrase slowly-growing epidemic, used by both, has been not-so-slowly appearing everywhere since then. Does that phrase accurately represent MERS to the world?

Yes, it does. If you have a look at the chart above, its been a steady increase ("blue mountain"), but despite the apparent steep slope of new cases, the steepest part of the mountain, extracted and plotted below, is in fact very linear. A steady but slow growth in cases. No exponential take off. No major deviations. So yes, there is an epidemic. And yes, it is slow. 156 (157 if 2nd Kuwaiti instance is confirmed) cases over 87 weeks in a country of 20,000,000+; a country that just hosted the biggest human gathering of the year (the Hajj) and a country which provides a launch point for around 18,000,000 travelers and a destination for almost as many



Click to enlarge.
A slowly growing outbreak of an emerging  coronavirus.
Cases have been accumulating worldwide but at a
linear rate since the week beginning April 7th.
Why the spike from this week? I'm not sure.
Feel free to use, just cite me and here.
But I think we need to be careful when throwing around the "E" word. An outbreak of an emerging virus may still be the best term to describe this chapter in the book of MERS. When someone asks on Twitter "to panic or not to panic"? (this was in reference to the latest MERS-map I posted) then I wonder if the correct message is being conveyed.
Another central message of the new LID paper was a no-brainer; well it was to me but perhaps I'm just too close to it all - in which case take this with a grain of salt.


I thought it was as obvious as the hump on a camel that where 1 case of a respiratory virus infection was detected, others were there to be found. After all, a virus needs us to survive - no us (which means no us actually harbouring infections, acting as a living incubator) then no more cases of the virus). Perhaps that's not obvious at all. Perhaps there is a lack of general understanding that our pathology laboratory systems do not test everyone with illness for even the "standard" endemic human respiratory viruses; that only those presenting to the right place, with the appropriate signs and symptoms, get a sample collected and get tested. This is apparently also true for MERS-CoV-which is by no means a standard virus. Do you go to your doctor if you feel mildly crook? Of course not - you go to work. What if you just have a fleeting headache, a stiff neck, feel a bit hot? Still going to work? Still going shopping? Still packing the kids off to school? Of course you are because we have these all the time and we have an immune system that does a wonderful job keeping it all mostly under control. Life goes on.
But you may be positive for a virus and you are a key part of the transmission chain. You are an incubator. A host.


So if routine testing is not geared towards finding out this extra information how do we find out what's going on in those who are not presenting with kidney failure or pneumonia; a relative small sliver of the population? Someone has to run a research study in which you enrol or get permission from people who are not very ill and sample them. Then you know something new about how widely the virus you are interested in is spread, for how long a person sheds it (if you sample the same person a few times during a month) and even how many other people get it (because all of a sudden your "contacts" become those of a less ill person and the numbers go up and you capture more of a picture of what's happening). So where are the research studies doing this?

When the illness is just some fleeting thing its no real problem. Especially when it's due to a virus we know all about and don't track for public health reasons (we track influenza virus positives, but the reality is you have to be sick enough to be tested in order to add to that pool of data). 



Click to enlarge.
A very exaggerated example of how failing to test mildly ill or asymptomatic
cases of infection in the community may confound our ability to make a link
between cases of severe illness leaving knowledge gaps. These gaps prevent our ability
to track spread of the pathogen and thus interrupt spread of disease.
Feel free to use, just cite me and here.

But if that virus is not yet in a textbook, not yet understood, not yet weighed and measured against the viruses we are more familiar with, emerges from an unknown place, is not considered endemic and is often notifiable, then not knowing this basic stuff becomes a major hole in our knowledge and our ability to respond appropriately. This is where we (still) are, 87-weeks after the first known MERS-CoV positive. Guessing (however educated) at what's happening by extrapolation and modelling.

I guess not everyone knows that for every time there is a noticeably ill person infected with a "respiratory virus", it's fair to assume that there will be at least 1 or other who gave it to them, got it from them or got it from the one who gave it to them and who are not as sick or even considered sick at all. For MERS-CoV, they are missed and thus we have no idea how the virus is spreading. Just models. But we can make mathematically supported guesses to back up gut instinct, fair assumptions and logic.


The hallmark of, and big problem with, the MERS outbreak (an epidemic mostly for the Kingdom of Saudi Arabia [KSA]), is that testing has been LIMITED to those who have pneumonia, or another severe disease, and their close contacts. Back in August Memish noted that surveillance was focused on those with pneumonia which was again noted by a WHO representative yesterday.


Why, why oh why not test more people? Why?! Is it because "it's too costly to prospectively test people by RT-PCR unless they are (very) ill"? It might be for some nations, but the KSA is not one of those. 


If you don't test others then you see these modelling publications arise. Idle hands and all that. Yes, it is great to have a model to support what many of us think to be true. And as Fisman and Tuite note in their editorial accompanying the LID article..



..inferences based on the best available data, even if those data are imperfect, allow decision makers to follow optimum courses of action based on what is known at a given point in time.

The question is, can decision-makers sign off on any actions if they don't have actual data? If those data are not forthcoming, how can we ever test the validity of the MERS models?

For now at least, I think we can agree that there is just too little testing to know enough to write more than a few chapters of the MERS-CoV textbook. A book for which we do have a table of contents. Many viruses have emerged before this one and they have each taught us what pages to skip ahead to. Unfortunately, we seem to have a recalcitrant author for 1 or 2 chapters.