VDU Quote for the night...

From:

Z.A. Memish et al. 

Screening for Middle East respiratory syndrome coronavirus infection in hospital patients and their healthcare worker and family contacts: a prospective descriptive study.

Clinical Microbiology and Infection, 2014.

"Surveillance studies also help in defining and monitoring transmission rates, case load, and epidemic risk assessment, and assist in instituting infection control measures with new diagnostic methods and treatments."

Pressure testing...

Comments in the recent ScienceInsider article (a great read by the way [1]) interview with Prof Christian Drosten got me to thinking.

What follows is a stream of consciousness around the need, or not, to expand laboratory testing capacity during times of an acute rise in cases such as during a viral cluster / outbreak / pandemic situation (COP; just made that up-it's not an official acronym or anything).

During a COP, the workload in a diagnostic virology/microbiology/pathology laboratory is dramatically increased. More samples, more often. And this is due to the testing of just 1 added virus. Often, the biggest impact on service delivery comes from the need to add a new test for this virus which may not have been part of any existing testing menu or panel; it adds to the number of tests already being run. In one major Australian laboratory, routine diagnostic testing for non-influenza-related diseases runs at ~1,000 tests/day.[2] In winter, Australia's peak season for influenza, this lab (Victorian Infectious Diseases Laboratory or VIDRL) would normally test ~100 samples per day for that 1 pathogen, but during the influenza A(H1N1)pdm09 pandemic, one day saw 1,401 tests done for it alone.[2] Impressively, these guys kept to their usual result turnaround time (TAT).

Such a response requires coping with extra paperwork, quality control, and the creation and implementation of new protocols, perhaps overcoming special specimen reception issues and specimen handling requirements. There may be delays in getting specimens to the lab and a need to enrol other (previously quality assured) COP assistance laboratories to cope with the load. Less urgent testing and research may be halted and even expanded lab space may be sought in adjoining areas. This all create some real impact. It can affect other results, it may impact on the TAT for a lab (although prior planning is aimed at coping with the strain of COPs and keeping the result TAT in check as happened in the example above). A COP strains nucleic acid extraction robots, centrifuges, bio-hazard safety cabinets, labelling machines, pipettes and thermal cyclers - all of which break down when you least need them to. Reagents may become rare and if not stockpiled could create a bottleneck in assay performance - basic PCR assay reagents may be hard to come by or slow to receive, especially during a global and/or sustained outbreak or pandemic. And very importantly, there is a real toll on staff and managers. Hours may be extended, tiredness and stress will set in and a shortage of expertise may be an issue for maintaining quality and TAT...and sanity

In other words, test results don't magically appear and diagnostic labs are nowhere near as automated as you might think.

All this adds up to a system that can reach its capacity and thereafter shows signs of stress. The influenza A(H1N1)pdm09 pandemic did this. Now we hear of that MERS-CoV may be creating a similar circumstance in the Kingdom of Saudi Arabia (KSA). Why is the KSA Central Laboratory, which does all the PCR testing for MERS-CoV, under such stress now? According to Prof Drosten, it's because of changes in the testing which may be a driving factor underpinning April's Jeddah surge of viral detection.

Something dramatic changed, and that is the case definition.

Prof. Drosten to ScienceInsider

This change led to a jump in testing from 459 samples for all of 2014 prior to the outbreak, to 4,629 in just 1 month. As the number of MERS-CoV tests being performed in each (daily) report of new cases is no longer part of the KSA Ministry of Health's (MOH) message, a thumbnail sketch is that 154 sample per day are being tested for that month (divided by 30 days). And then there was this comment..

"The question of whether there is a mild, short-lived infection in some people is scientifically interesting. But in cities like Jeddah, it is bringing the health system close to collapse. That is the big problem. So many samples are being tested that the lab capacity won’t suffice for the real cases."

Prof. Drosten to ScienceInsider

An entirely fictional map of MERS-CoV spread including
severely ill, mild/moderately ill and prodromal /
asymptomatic infections. Simply intended to be
something to think about when discussing the
impacts of limiting PCR testing. Reduced PCR
testing should not happen until until we know which
parts of this map are real, and which are a load of rubbish.

With this background and these comments in mind I have some thoughts and questions...

1. I know almost nothing about the KSA's pathology laboratory testing capacity generally nor its approach to respiratory virus testing in particular. I do know that the KSA is are a country of around 29 million people while Australia has around 23 million. I refer to the numbers above when I say that 4,629 samples in a month, for what has become an epidemic that seems to have exposed major flaws in infection control across multiple hospitals around the west, south and central regions of a wealthy country, should not be threatening the KSA's testing capacity unless it did not exist in the first place.
2. Why wouldn't pathology testing which is robustly designed to cope with a worse-case-pandemic, not exist in the KSA? I don't know. Does testing exist for standard virus screening and if so what sort of throughput is the norm? The KSA healthcare systems seems to be laden with western-influenced medicine, and with that influence comes our compulsive need to create protocols and preparedness plans and to learn for the misfortune of others. The WHO have all this sort of information publicly available and always seem available for a chat.
3. The reality is, and I am not on the ground to see whether this is a real factor in the KSA, laboratory capacity needs to be such that it can cope with a surge in cases such as that during a COP. It also needs to manage other endemic respiratory virus testing and whatever is coming next. It seems highly likely to me that the same at-risk older male population with kidney and heart disease, diabetes and obesity issues that get hit so hard by MERS, is also suffering badly from influenza and other viral infections. Back in August we heard about additional laboratories coming on line. It looks like they may not have. They need to.
4. Am I especially naive (probably) to expect wealthy countries to make sure something as important as pathology testing is not in danger of falling over when it's particularly needed? We expect our electricity to be quickly reconnected after a storm, out SUVs to be easily refuelled no matter what wars or disaster befall the worlds, we take for granted that water is just there and we'd riot if our shop were not stocked with food 24/7. Why would testing your population to make sure you have a real-time knowledge of the pathogens infecting them, not be given an equal measure of attention and support? Especially if that pathogen has never been seen before, is transmitting without your understanding and is killing 1:4 of those it infects?
5. Prof. Drosten noted that he has been working to get good MERS-CoV antibody testing in place within the KSA to get a better idea of how widespread prior exposures to MERS-CoV is. That will be a very helpful piece of knowledge to have. But it will not tell the MOH what is happening now in Hospital X (an apt name since we no longer know names of the hospitals where cases are being treated; that dropped off the new MOH messaging format last night). We're not even sure MERS-CoV antibodies are produced if the PCR-positive person only had a mild or asymptomatic case. PCR testing must remain in place until the MOH or whomever it looks to for advice, can be sure they have seen all the faces of MERS and the MERS-CoV. We're some way off seeing that yet I believe.  Don't get me wrong - an antibody test is great and we should roll it out alongside PCR. But in context - it will tell us information about the status of the KSA population in terms of how many have been exposed to MERS-CoV. And then it will have done its job as a research tool. Routinely, we need to test with the gold standard; PCR. And I think we should keep testing widely. 
6. Prof Drosten also suggested that instead of continued PCR testing of contacts (the source of asymptomatic cases presumably), the KSA should consider a home isolation approach. Would that be  for up to 2-weeks, away from work, school - away from family too? Seems like a lot of hassle and disruption for the sake of a PCR test. Perhaps a shorter period once we know more about the dynamics and shedding during the diseases prodrome or from asymptomatic people. That will require PCR to define a person was initially MERS-CoV positive in order to study whether virus is shed.
7. Let's also keep in mind that antibody testing is labour-intensive too. Perhaps not as intensive as PCR, but it would still increase the workload on a pathology laboratory.  It's a new tool not a better one.
8. Why do I think the KSA should keep testing widely? Because if we don't we might be missing mild and asymptomatic or prodromal cases which may (and we have no data to support the argument in either direction right now, so its much better to be safe and test as the World Health Organization advise) contribute to the spread of MERS disease. Who knows how much virus an already old ill male needs to become severely ill? Perhaps much less than a healthy young nurse with lots of previous exposures to other viruses, including some that may provide cross-protective immunity I suspect. 
9. If the KSA had not switched gear and accelerated into more testing, we would still only know the face of MERS that is pneumonia and death. It is clearly a lot more than that-as are all respiratory viruses. It would be a great shame in my opinion, to do things the way they were done with SARS, just...because. We always need to look afresh with the knowledge and tech we have to hand on the day.

Now more than ever with new measures being instigated to educate the KSA public (a bit more anyway), reduce camel exposures (although it's clear many don't see a link to camels as justified) and improve hospital infection control (too late for the majority of MERS cases that seem to have occurred in linkage with healthcare facility outbreaks) and hospital triage of MERS cases, testing efforts must not wane.

And while that goes on in the background, it really is past time to sort out some transmission details. How is the virus spread (a) from and between camels and (b) to and between people? These are fundamental questions and all risk reduction hinges on their answers. 

At least now that we know the virus hasn't changed, we shouldn't be seeing any more cases during the upcoming multi-million person Hajj pilgrimage, than we saw last year. Right? Last month was all about an infection prevention and control breakdown that can be fixed before October. Yes? And the few instances of Umrah pilgrims that seem to be popping up positive this year that we didn't see in 2013 and the bunch of single export cases? Just increased testing? Yup. Some of that even kinda fits in with what I wrote about Umrah 2013. 

Oh look. 10 new cases tonight, just like on 2013. Oh wait. No it wasn't like tat in 2013. We didn't have any 10-detections/day days in 2012 or 2013. Guess these will be because of all the pesky asymptomatic people? Let's see...ICU, hospitalised, ICU, symptomatic but home isolated, ICU, ICU, asymptomatic, ICU, asymptomatic, hospitalised oh and in two most likely unlinkable previous cases: death, death. 2 out of 10 with no symptoms. 

Definitely keep up the testing guys. MERS isn't SARS but then 2014 isn't 2013 either.

Sources...

1. http://news.sciencemag.org/health/2014/05/mers-virologists-view-saudi-arabia
2. Reality Check of Laboratory Service Effectiveness during Pandemic (H1N1) 2009, Victoria, Australia | Emerging Infectious Diseases. 2001. 17(6):963-
   http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3358210/pdf/10-1747_finalS.pdf
3. http://www.nccid.ca/files/Evidence_Reviews/NCCID_H1N1_impact_04.pdf

If this is what MERS-CoV detections look like with more testing...what is the "normal" community level of virus?? [UPDATED]

The bar along the top depict how much time passed between each 100 cases.
Click on chart to enlarge.

For a virus that is chugging along without the aid of any new genetic changes, and perhaps showing up more often (a) because of enhanced testing and/or (b) because of a large-scale breakdown in infection prevention and control (IPC), this curve sure does depict the possibility that we had no idea how much MERS-CoV was transmitting among the population. Still a poor transmitter compared to an influenzavirus, because we have seen a few larger MERS-CoV studies than show few to no MERS-CoV positives, but still more people positive than we thought.

Can we really lay this rate of climb at the feet of poor IPC alone? Wouldn't that also mean that every other respiratory virus would do this too? Perhaps not if MERS-CoV was the only one capable of causing acute pneumonia. It isn't. So shouldn't all hospitals always be full of acute pneumonia and respiratory disease among older males with comorbidities? These are modern hospitals after all. Also modern doctors with great training and skills acquired from all over the world. Perhaps poor IPC plus enhanced testing is an option? Maybe. Probably most likely when combined with an outbreak that starts in April for an unknown reason.

What about this option, which focusses on testing alone? Better levels of testing are at last showing it like it really is in the Arabian peninsula? This options proposes that we've been underestimating the ability of MERS-CoV to travel from person-to-person, all along. An underestimation driven by testing only the "tip of the iceberg" of disease and just watching the rest of the iceberg from a distance? What if severe disease is only found in already ill older males and most (granted, not all) of the rest get milder or unnoticeable disease but do get infected? Yes, in the past 2 weeks 5,000 samples have been tested in the Kingdom of Saudi Arabia (KSA) to yield ~140 MERS-CoV detections (~3%). After the 2012 Hajj, 154 pilgrims were tested by Gautret et al and, despite a high proportion with respiratory symptoms (83%), none were found positive for MERS-CoV. At 3% in Jeddah, testing of the Hajjis should have yielded ~4 detections among this cohort if the distribution of MERS-CoV was at this level all the time. In screening 5,065 cases and their contacts (family and healthcare workers) over a year from 1st-Oct-2012, Memish et al reported 106 detections (~2%) and no significant rise in case detection rates over that year. So a very similar proportion of positive cases, and both Jeddah and this larger study have similar numbers and a roughly similarly broad population being tested. Of course, even endemic human CoVs are not always detectable every year at the same site. Some have a biennial periodicity.

So perhaps 2-3% prevalence, similar to endemic CoVs, is the magic number for MERS-CoV positivity? Which just leaves the question, why weren't more Hajjis positive last year? Or was that just a testing thing too?

Just wondering out loud here.

Ebola virus disease and lab testing...

Virology Down Under's latest Ebola virus case case chart.
Click on chart to enlarge.

Maia Majumder has posted a nice concise comment on her blog. In her latest post [1] she notes that we shouldn't be too surprised that the number of Ebola virus disease (EVD) cases with a lab confirmation (conf) represent a relatively low proportion of the total cases we hear about. 

Currently (see the chart above), 35.3% are lab confirmed. That's 59 confirmed among 167 cases; the remainder are suspected [susp] or probable [prob] cases.[3]

What might contribute to the speed of laboratory confirmation in this and other EVD outbreaks? 

Some thoughts below:

• Obtaining a specimen. If a body has already been hidden, buried or otherwise disposed off before a sample can be collected. Sampling may have been refused by next of kin-although I am not at all sure if that is a "thing"  during an EVD outbreak
• The need to work under enhanced safety conditions to prevent laboratory-acquired infections. BSL4/PC4 not strictly available to the field labs (although they are setup to work with those pathogens; see Tweet below), but increased care and awareness still slows down the diagnostic process compared to testing for a much less fatal virus
• The generally tough conditions for doing precise and careful lab work in a mobile laboratory; work that is often resource-, temperature- and power-sensitive not to mention fiddly and in need of well-controlled experimental conditions
• Distance from the site of collection to qualified lab and the quality of sample once it reaches that lab. A sample that sat around in the sun or was accidentally frozen, lost, broken, sent to the wrong place, may be falsely or weakly negative requiring further testing
• The case is positive for a different virus but one that causes similar signs and symptoms. This may also require additional testing to identify. Other virus testing may be run in parallel..or may not
• You could argue that previous outbreaks used older and often much slower diagnostic methods. That's true, if you compare them side-by-side in a results race. In practice, PCR-based testing comes with lots of extra "bits" that can slow down the production of a final result. The process is still faster than things used to be, quite possible more  sensitive too, but still not as fast as we'd all like. Apples and oranges though.

#Ebola samples now tested locally in #Conakry and Guéckédou, #Guinea http://t.co/oOYsd1PPGW [PHOTOS] #AskEbola pic.twitter.com/2UfehauCs7
— WHO (@WHO) April 5, 2014

What defines a suspected case requiring testing anyway? 

Pretty much the same things that define this for any outbreak; a suspected case is a person with the appropriate signs and symptoms of disease, who was in the right place at the right time to have come into contact with a known infected human or animal in such a way that they may have exposed themselves to virus, but they have not yet received a lab confirmation that they have that virus. It may be that a case never receives that confirmation because of a lack of positive specimens (don't have specimen or cannot get a positive result) in which case the person becomes a probable case if they meet the clinical criteria but cannot be confirmed. 

Why would a sample not be collected? 

As noted above, perhaps the next of kin did not allow samples to be collected, perhaps the body was disposed of before sampling could be achieved or perhaps the lab testing failed. To safeguard against the latter, PCR-based testing (not the only method) usually involves multiple assays, running replicates of each sample, and using several assays, each preferably targeted to a spatially different region of the viral genome to overcome the negative impact of any genetic changes in relying on a single site. Such viral genetic change may be an issue during a new outbreak. We haven't seen much by way of sequence analysis from any viral detections to date, but very early on in this outbreak the species was confirmed using genetic sequence determination, to be a strain of the species Zaire ebolavirus.

The numbers are constantly changing.

After all that, even a probable case may still get be discarded after lab test results are in; it may have been a suitably relevant disease, but caused by infection with a completely different virus.

While I think many of us understand that the numbers do change, I also think some of the interest we have in wanting to see them is to understand which way the trends are changing; up, down, steep, flat etc. There has been a fair bit of cautioning about the numbers. In my own defence, these numbers are real. They are collected by people on the ground. They are a much better metric to watch, changeable or not, than the many headlines and blogs and Tweets that may be more aimed at attracting readers and followers, or just be ill- or uninformed.

So the numbers change. What does that mean? As it stands, the Sierra Leone cases have now been taken off the Ebola tally because they were confirmed as haemorrhagic fevers due to a completely different virus; Lassa virus. A suspected EVD case in a child tested negative in Ghana. 2/6 suspected cases from Mali have also tested negative for the Zaire ebolavirus. The Liberian hunter thought to be an isolated EVD acquisition [8,9,10] not linked to Guinea, has now tested negative for the virus. So the numbers change quickly. That's your proof and it confirms what WHO's Gregory Haertl has been saying since Day 1 of this outbreak. These changes have effects too.

The fatal case percentage may rise despite more cases testing negative.

Not as strange as it sounds.

If the number of susp/prob cases drops as some are discarded because the lab confirms they are not EVD, the proportion of cases that are confirmed and died due to EVD will "look" larger-it will be a bigger percentage. The proportion of fatal cases currently sits around 63% of all susp/prob/conf cases now (up from a lowest point of 59%, down from a high of 72%). If the denominator (total susp/prob) cases should shrink while the numerator (fatal EVD cases) remains steady, or grows, the ratio will grow. Be prepared for that and the accompanying headlines or poorly informed Tweets and comments that will scream "the virus is mutating" blah blah blah. It probably isn't. It probably won't. But you may not get that message from using Google alone (try the links below and work your way up).

This EVD outbreak is proving especially challenging.

The term "challenging" seems to have become an agreeable descriptive for both the WHO and MSF, at last, as of yesterday's WHO virtual press conference[5]. 

The challenges that differentiate this Ebola outbreak from previous mostly seem to be about the wide spread of cases around the countries of both Guinea and Liberia, complicated by the presence of other pathogens that cause clinically similar diseases. More usual problems for tracking, identifying and confirming EVD cases are listed above including working under the requirements of enhanced safety and the need to bring in many essential resources. Careful and accurate confirmation of cases by the lab is a time-consuming process but one that must be given that time in order to ensure it gets the right result. False-negative results or lab-acquired infections would be a very bad outcome at any time but especially if resulting from an unnecessarily rushed testing process. False-positive results have an arguably larger negative impact on the entire situation. Timeliness is a very subjective thing. But lab confirmation is most definitely not like making a cup of coffee.

Can we see the forest for the trees yet?

The most recent EVD susp/prob/conf cases became symptomatic on 06-April-14, but no new healthcare workers were among them and some cases are now being discharged .[4] Some good news there.

We're obviously not out of the woods yet (pardon the pun) in terms of transmission chains. The WHO suggests it will be "some months" before we stop seeing cases. But the recent WHO virtual media conference stressed that while EVD is a serious disease it is one that can be controlled and the risk of infection is low, when the right precautions are in place.[5]

Dr Fukuda: We expect to continue seeing #Ebola cases for some months, to go through several incubation periods. Premature to say decreasing
— WHO (@WHO) April 8, 2014

See the latest WHO-AFRO Ebola in Western Africa Situation Update also. It's got totals and charts!! Bloomberg quicktake webpage [6] and the US CDC webpages [7] have lots of digestible information too.

References...

1. #Ebola2014: On the Topic of Lab-Confirmation
   http://maimunamajumder.wordpress.com/2014/04/08/ebola2014-on-the-topic-of-lab-confirmation/
2. WHO-AFRO Ebola virus disease (EVD), West Africa Situation Report 07-Apr-14.
   http://www.afro.who.int/en/clusters-a-programmes/dpc/epidemic-a-pandemic-alert-and-response/outbreak-news/4089-dashboard-ebola-virus-disease-in-west-africa-07-april-2014.html
3. WHO GAR DON Ebola virus disease (EVD), West Africa Update 07-Apr-14
   http://www.who.int/csr/don/2014_04_07_ebola/en/
4. SUCCESSES AND CHALLENGES IN RESPONSE TO GUINEA EBOLA EPIDEMIC
   Médecins Sans Frontières Press Release 08-Apr-2014
   http://www.msf.org.au/media-room/press-releases/press-release/article/successes-and-challenges-in-response-to-guinea-ebola-epidemic.html
5. Audio file for WHO virtual press Conference
   http://terrance.who.int/mediacentre/presser/WHO-RUSH_Ebola_outbreak_Guinea_presser_08APR2014.mp3
6. Bloomberg's QuickTake on Ebola
   http://www.bloomberg.com/quicktake/ebola/
7. The US Centers for Disease Control and Prevention on Ebola in West Africa, 2014
   http://www.cdc.gov/vhf/ebola/outbreaks/guinea/
8. Liberia reports suspected Ebola outbreak unconnected to Guinea
   http://news.yahoo.com/liberia-reports-suspected-ebola-outbreak-unconnected-guinea-130714958.html
9. LIBERIA: Ebola Deaths Rise In Liberia, Health Minister Confirms
   http://www.gnnliberia.com/articles/2014/04/05/liberia-ebola-deaths-rise-liberia-health-minister-confirms
10. Liberia: An isolated Ebola case
    http://crofsblogs.typepad.com/h5n1/2014/04/liberia-an-isolated-ebola-case.html

Google Flu Trends: not so perfectly predictive?

I'm no expert at the algorithms that go into the search giant's Google Flu Trends (GFT) predictive website so take what follows as a very superficial opinion. It does not surprise me at all that a recent paper in Science [1], backing up previous chatter on this subject [2], finds GFT is is not very accurate. Specifically, it has been overestimating peak influenza levels compared to more traditional laboratory-confirmed cases (itself only a subset of all cases) and influenza-like illness presentations to Doctors (a non-specific method of trying to identify influenza from a swarm of other ILI-capable viruses). 

A note: this recent paper is more a look at big data and whether it deserves our complete trust yet (it doesn't, is the message) than it is an analysis of how best to predict influenza virus activity in the future.

It would be fantastic if we could have a predictive system that could work around the need for actual testing of sampled people and give us an informed guess as to what flu was doing, how long it would be doing it, how severely it would do it and when it might start and stop doing it...I just don't have a lot of faith in predictive things like this. Perhaps I've just entered into a grumpy middle-aged male phase of my life....but I think that if we want to find out what's happening, we don't need to look too much beyond simply (not so simple when it comes to lab capacity and funding of course) upping the level of testing and typing that we do.

Even now, the current situation in Queensland of a 2-fold increase in influenza virus notifications compared to the mean of the past 5-years does not really show up so clearely on GFT.


Given that so many variables will contribute to a person's choice to search for "flu" (or whatever related text GFT includes in its algorithms), it makes perfect sense that a website showing flu activity in your area that is based on that component of the results, will be an over-estimate, especially during the peak times of flu activity. Why then? Imagine the impact on search when the media is most active in trying to get your pageviews using headline banners with "killer flu" or "early flu season" in them. People don't just chat over the back fence in response to those headlines any more, they go looking to the internet to provide their answers, news and sometimes poorly communicated facts. This will not just indicate that the have the flu, it will reflect concern that they may get it at some point in the future.

GFT also taps "real" flu data from real testing labs and Doctors clinics. This means its performance is probably not "off the rails" wrong, just overly influenced by non-infectious factors at peak times.

Are the inflated results positively affecting flu vaccine uptake I wonder? That would be a good thing. Might even have an impact on the size of the peak season.

Of course, no one knows what the actual numbers of flu cases in the community are; because flu is not always a serious disease that leads us to get a sampled collected and tested. The serious disease gets to a hospital and does get tested. these get added to notifications. Sure, influenza virus can cause a more serious outcome than many other respiratory virus infections, especially in certain groups, including death on occasion. There are also many mild infections that fly "under the radar". Those numbers won't be accounted for anywhere except through modelling. Perhaps the overestimate isn't that much of an overestimate; very hard to actually know that.

It's that damn iceberg tip again. 

References...

1. http://www.sciencemag.org/content/343/6176/1203.summary?rss=1
2. http://www.nature.com/news/when-google-got-flu-wrong-1.12413

Neither market nor farm poultry all that positive for H7N9; songbirds the culprit...?

Following on from yesterday's post, "If not poultry then what?", I thought it worth noting the impressive numbers from the Chinese Ministry of Agriculture.

From 2013:

• 1,630,000 poultry and environmental samples tested
• 88 POS; all from live bird markets
• None from poultry farms

From 2014, to date:

• 33,400 poultry and environmental samples
• 8 H7N9 POS; all from live bird markets
• None from poultry farms

The other alternative to answer the question in my heading; the testing methods are at fault. 

No detail of what approach has been used to obtain these numbers in the links below. Viral culture and serology with some PCR have been noted before. I'd wager culture yields chicken scratchings compared to PCR for detecting virus in he wild; but serology has successfully been a pillar upon which animal testing rests. So that's why the numbers above are such a quandary for the epidemiologist who reads about the high frequency of links between human disease and exposure to poultry.

It would be nice to see some technical papers on antibody test testing (development and validation) at some point. If only to reassure everyone that the testing methods are doing what testing methods should be doing.

See #3 below for influenza PCR discussion at WHO.

Sources..

1. http://english.agri.gov.cn/news/dqnf/201304/t20130427_19537.htm
2. http://english.cntv.cn/program/china24/20140130/100709.shtml
3. http://www.who.int/influenza/gisrs_laboratory/report_2012pcrwg5thmeeting.pdf

H5N1 case in Canada had been diagnosed with pneumonia...testing at the source would have been helpful

And now, from a fantastically detailed post onto ProMED by Fonseca and colleagues, we see that the H5N1 case was diagnosed with pneumonia.

On 28-Dec, the patient presented to a local emergency department.

"A chest X-ray and CT scan revealed a right apical infiltrate. A diagnosis of pneumonia was made; the patient was prescribed levofloxacin and discharged home."

One sad point made in the ProMED post which supports the need for constant viral vigilance the world over, coupled with the dissemination of those surveillance data, so that patient management anywhere in the world can be armed with the best possible decision-making information...

"The index of suspicion was low as travel was to an area in China where there have been no recent reports of the circulation of this virus, and coupled with no obvious exposure to poultry, the diagnostic work-up and consideration for A(H5N1) infection was very low"

As a recent J Virology article by Yu and colleagues highlights, when a sensitive testing method like the polymerase chain reaction (PCR; in this case RT-PCR because influenza viruses all have an RNA genome, not a DNA one) is applied to the search for a virus, it yields the kind of data that can:

1. Explain from where a virus emerges
2. Inform the search for disease aetiology - where are human cases getting infected from and if a zoonotic infection (from animals to humans), which animal(s) is the culprit?
3. Alert the world to any risks of infection when travelling to a certain area(s)
4. Allow the local health departments to mitigate the risk of their population acquiring infection by instigating controls (like live bird market closures). This has implications for the world since respiratory viruses have the potential (thankfully not realized for H7N9 or H5N1 to date) to spread more rapidly and efficiently that blood-borne or mosquito-borne or sexually transmitted viruses.
5. Permit understanding of how widespread (over what geographic area is it detected) a novel or emerging virus may be and how entrenched (is the same site repeatedly positive) it is

Not doing such testing, or using less sensitive methods will not yield this information. 

In Yu's study, testing of 12 poultry markets, mostly urban, and local farms linked to 10 human infections in Hangzhou, Zhejiang province around 4th to 20th April 2013 yielded signs of H9N2, H7N9 and/or H5N1 viruses in all markets. Poultry were often positive for H7N9 and H9N2 (this finding from individual RT-PCRs was confirmed using next generation sequencing), whereas human specimens were not. These levels hadn't been turned up when 899,000 bird were tested in 2013 using (perhaps) less sensitive methods.

I think with influenza, it may be safer to presume its everywhere until that presumption can be discounted. Clearly the conditions for influenza viruses to swap gene segments and sort themselves into new subtypes and variants are commonplace and frequent; these aren't just chance occurrences of different birds passing in the night via overlapping flyways. These feathered vectors are co-infected by 2 or more viruses at a time. Luck and the constraints of viral fitness are presumably the only things keeping H7N1, H5N9, H7N2 cases from dialing up in humans? What seems to be lacking is more molecular testing at the farms supplying the markets. Not just in Zhejiang, but all over the region.

As the authors noted, 100,000s of people visit these live bird markets each day and very few influenza cases seem to be due to them. Long may that last. But it's a tinderbox for which matches are already being struck; if the viruses should bud of that one-in-a-million variant that is enabled to readily spread from person-to-person, whooshka. 

More testing guys, keep testing.

Hajj pilgrims return around the world...is lab testing happening?

Media and Ministry reports are filtering in from many different countries that their Hajj pilgrims have been safely returning from their pilgrimage. 

To date there have been no reports of Middle East respiratory syndrome (MERS) disease in any pilgrim. 

I presume this is all observational diagnosis? It would be very interesting to read whether any actual laboratory testing is occurring in any of these States. If it is, are the pilgrims PCR-negative?

What observation alone cannot tell us is whether a pilgrim infected but not showing signs. 

Self-reporting of mild disease without overt signs, can be problematic and may bias away from capturing all cases in the absence of laboratory testing. If cases do enter a State "under the radar" they may still shed virus to others in their new locale. Some of those others may be older males with comorbidities; the MERS coronavirus's (MERS-CoV) highest impact population. Of course we don't really know if mild and asymptomatic cases can transmit effectively. We might if their was more widespread testing. Since we haven't seen that level of testing coming from the site of most MERS-CoV infections, the Kingdom of Saudi Arabia, perhaps there is an opportunity for other States to step up and test not just returning pilgrims but their families and other contacts and see whether their upper respiratory tract's are free of MERS-CoV RNA?

These media reports also don't tell us what definitions are used by each State to define a pilgrim as being free of MERS-free. It may be the absence of any sign of any respiratory disease, or it be just absence of severe signs and symptoms, or perhaps a combination of signs and symptoms e.g. fever+cough or cough+difficulty breathing.

Too many knowledge gaps. 

One thing's for sure. The headlines are only scratching the surface.

The tip of the VirusBerg...

The VirusBerg.
Click to enlarge. If there are more cases with mild/less
 symptomatic disease then the denominator grows
and the proportion of deaths to survivors, the PFC,
shrinks. At the start of an outbreak or emergence of
a new pathogen, it is usually the most severely ill or
those ill with a specifically identifiable disease
that attract the attention of specimen collectors &
testers. Later in the event as resources are freed up
and panic reduces, testing of other population
groups may find more cases of less severe disease.
 This may not eventuate though and the PFC may
 remain high confirming a serious threat to human
health. Feel free to use graphic. Cite this blog
and Ian M. Mackay.

I've talked about the Denominator problem over on my "main" site, Virology Down Under's H7N9 page, but I thought it worth revisiting for 2 reasons: (1) our understanding of Middle East respiratory syndrome (MERS-CoV)  fatalities may still affected by a possible bias in our understanding of MERS-CoV transmission because of a continued focus on testing just the "tip" of the VirusBerg and (2), I made a new graphic that I like more than the old one!

We see a high PFC for MERS for 2 reasons that spring to my mind:

1. There are no, or very few, mild cases of MERS. Cases are usually severe and most are being caught. We know there are some mild cases and we know that contacts of confirmed cases are infrequently MERS-CoV positive. I'm assuming contacts are tested with the same rigour as those in hospital due severe MERS (e.g. lower airway sampling with follow-up testing). 
2. Although we don't see much transmission from cases to contacts we do see some and this may be enough to maintain a small number of community transmission events. The second reason then is that we are only testing cases with severe disease, many of whom die as a result of complications due to the initial infection. Prospective screening of a sample of the well general community and those with "common colds" might identify more cases that would reduce the PFC.

4.8-million Umrah pilgrims free of MERS-CoV...?

According to an Arab News report, the Kingdom of Saudi Arabia's (KSA) Health Minister Abdullah Al-Rabeeah, has said that 4,800,000 pilgrims visited  to perform Umrah this year, and not one left having had a Middle East respiratory syndrome coronavirus (MERS-CoV) infection. 

That certainly suggests that the virus is hard to catch and that its not transmitting stealthily.

Wait...it doesn't mean either of those things. 

What it means, in case you were to misinterpret the headline, is that the disease, MERS (as opposed to the virus, MERS-CoV), was not diagnosed in those people. 

Wait, no, it means that a disease severe enough to be put on the radar for MERS testing, did not occur among the 4,800,000 pilgrims. Just a reminder from the World Health Organisation of what their interim definition of a probable MERS-CoV case looks like: 
excerpted form the 3rd July 2013 version

1. A person with a febrile acute respiratory illness with clinical, radiological, or histopathological evidence of pulmonary parenchymal disease (e.g. pneumonia or Acute Respiratory Distress Syndrome)
   AND
   Testing for MERS-CoV is unavailable or negative on a single inadequate specimen
   AND
   The patient has a direct epidemiologic-link with a confirmed MERS-CoV case
2. A person with a febrile acute respiratory illness with clinical, radiological, or histopathological evidence of pulmonary parenchymal disease (e.g. pneumonia or Acute Respiratory Distress Syndrome)
   AND
   An inconclusive MERS-CoV laboratory test (that is, a positive screening test without confirmation)
   AND
   A resident of or traveler to Middle Eastern countries where MERS-CoV virus is believed to be circulating in the 14 days before onset of illness.
3. A person with an acute febrile respiratory illness of any severity
   AND
   An inconclusive MERS-CoV laboratory test (that is, a positive screening test without confirmation)
   AND
   The patient has a direct epidemiologic-link with a confirmed MERS-CoV case.

So really, only #3 would account for any MERS-CoV infection that wasn't at the severe end of the clinical scale (the tip of the iceberg as we all like to refer to it). But even then, a case has to be linked to another MERS-CoV case. 

In other words, less obvious cases - those infections that may look like a standard influenza-like illness (ILI), or be asymptomatic (both have occurred win MERS-CoV positive people) - would not get tested and we have no way of saying that these people were MERS-CoV free.

I'm sure that many of those 4,800,000 pilgrims had some respiratory symptoms during their time in the KSA - we know from research papers that the 200 or so endemic respiratory viruses continue to circulate among pilgrims in the KSA during large gatherings.

It would be very reassuring to know how many cases of mild or moderate ILI respiratory infections were seen by a Doctor in the KSA and how many were tested for MERS-CoV by RT-PCR. A prospective study like that by Gautret et al which did not sample on the basis of signs and symptoms, augmented with some serology testing to show recent or past MERS-CoV infection but conducted by/at/for the KSA, would be a great example. It would even pick up asymptomatic cases. Even without the serology component that would be a valuable good study.

I guess we'll just have to wait for the research paper.