I just found this article on Bloomberg.com regarding the H1N1 Swine Flu; the article is entitled: "Swine Flu Is Spreading Wider Than Official Data Show (Update2)", and this article is nothing more than scare tactics and sensationalist journalism at its worst. It might take a trained eye to catch some of this stuff and interpret it, but that's what I'm here for. In any case the first little bit of the article reads:

Swine flu is spreading more widely than official figures indicate, with outbreaks in Europe and Asia showing it's gained a foothold in at least three regions.

This first statement is meant to frighten you, meant to lead you in and make you believe that the swine flu is a serious concern - or at least more or a concern than the typical seasonal flu. The italicized portion is the real scare tactic... in reality it's actually quite an innocuous statement, and is explained in the very next paragraph/statement:

One in 20 cases is being officially reported in the U.S., meaning more than 100,000 people have probably been infected nationwide with the new H1N1 flu strain, according to the Centers for Disease Control and Prevention.

Sounds scary, doesn't it? In actuality, it's not only not a bad thing - it is in fact a good thing, and here is why. In my previous swine flu update, I provided some statistics regarding Case Fatality Ratios (CFR's) for not only past 'serious' flu outbreaks, but also for the seasonal flu, and this current swine flu outbreak, rather than reproducing these statistics here it is easier to simply quote this other post:

...the CFR is the best estimate or description of a disease's virulence. Estimates for the CFR of the seasonal flu are approximately 0.1%. The 1918 Spanish Flu Pandemic has been estimated to have a CFR ranging from 2% to 20%, depending on which sources you check, but is most commonly given as 2.5%. This particular virus, depending on which statistics you look at. Considering only the number of deaths per infected individual, the USA, Mexico, and worldwide have CFR's of 0.110%, 1.97%, and 0.78% respectively. These CFR's are likely overestimated as well; it's reasonable to assume that the actual number of cases is higher – probably significantly higher – than the reported number of cases.

As the quote above indictates, the CFR's that I have calculated are actually likely to be overestimated by at least an order of magnitude, or as suggested above as much as 20 times. If I recalculate CFR's based on an underestimate of 20 fold, the numbers as of today for the USA, Mexico, and the entire globe are 0.00739%, 0.09635%, and 0.036525%, respectively. In the worst case scenario, Mexico, based on the most recent swine flu update from the WHO, the approximate CFR is 0.09635%, under the value typically estimated for the seasonal flu. If you read the article a bit further, the prognosis is even more hopeful:

In the U.K., the virus may be 300 times more widespread than health authorities have said, the Independent on Sunday reported yesterday.

Or based on the most recent numbers of 137 cases and 0 reported deaths, the actual and estimated CFR's are zero, given the underestimation of the case number, the number is less than in the US or Mexico when each of those countries had a comparable number of cases 0.709% and 5.769% for the US and Mexico, respectively. Even more good news:

Japan, which has reported the most cases in Asia, began reopening schools at the weekend after health officials said serious medical complications had not emerged in those infected.

The statement that follows attempts to frighten the reader once again:

The virus is now spreading in the community in Australia, Jim Bishop, the nation's chief medical officer, said yesterday. "I think we will see the number rise," Bishop told Australian Broadcasting Corp. radio today after confirming the nation's 17th case and saying test results are pending on 41 others.

The flu virus spreads, and as it spreads, the number of cases rise; this is what's expected of the flu virus, or for the most part, any easily transmitted virus. As the virus makes its way around the planet, more people are infected. Nothing scary about that - or at least nothing more scary than the seasonal flu. The statement that truly drives home the point that - thus far - there is nothing to fear from the dreaded H1N1 swine flu:

Most of those infected experience an illness similar to that of seasonal flu. The main difference is that the new H1N1 strain is persisting outside the Northern Hemisphere winter.

For the most part the rest of the article simply cites statistics about the virus - all of which suggest that this particular virus is of no more concern than the typical seasonal flu.

The swine flu - for the most part - seems to have fallen off the radar of the mainstream news. There still several things about the swine flu that are not known. In particular, it is not known how severe this particular flu outbreak will be with respect to illness and death relative to other influenza viruses.

According to the WHO, disease surveillance authorities in Mexico began noticing cases of a disease initially described as a mild influenza like illness starting on March 18. The number of cases has risen steadily since then, and again, according the the WHO, there are now 10243 cases and 80 deaths in 41 countries worldwide. The large majority of these, 5469 and 6 deaths, have occurred within the United States; the second largest number of cases, 3648 and 72 deaths, have occurred within the epicenter of the disease - Mexico. That there are more cases of flu in the USA relative to Mexico, the epicenter of the disease, most probably speaks to the relative reaction of Mexico as compared to the USA.

I thought it might be noteworthy to plot the progress of the disease to illustrate this idea. Below is a semi-log plot that follows the number of flu cases, beginning on April 29, 2009, the day that WHO first began posting any meaningful data. [caption id="attachment_63" align="aligncenter" width="497" caption="Semi-log plot: Number of Flu Cases vs. Date "][/caption]

Perhaps the most noteworthy item concerning this data is the fact that the number of cases in the USA surpassed the number of cases in Mexico relatively early on in the outbreak. The disease was handled completely differently in the two countries. In the USA, very little action was taken - though some action was taken in some isolated places. For example, 858 schools were closed throughout Texas, affecting more than half a million students; a number of schools were closed throughout California, and NYC has closed at least 16 schools.

The response in Mexico has been somewhat different; Mexico reacted to the virus by closing businesses, restaurants and schools, canceled concerts and sporting events, and advised citizenry to stay indoors as much as possible for five days. This strategy - logically - is thought to be largely responsible for keeping swine flu under control and limiting spread through Mexico. In other words the Mexican response was much more logically consistent with actually controlling an outbreak than that of the USA; and the data clearly demonstrate this.

By any standard the US response to this unknown, novel, out-of-season, potential pandemic has been lackluster and nonchalant.Whether or not a less passive approach was necessary remains to be seen. The virulence of any given disease is defined by it's case fatality ratio (CFR); a higher CFR is indicative of a more virulent virus. The tricky thing is how CFR is defined. Ideally it should be defined as number of deaths, per affected individual, however how those numbers is defined is also often tricky. In any case, the CFR is the best estimate or description of a disease's virulence. Estimates for the CFR of the seasonal flu are approximately 0.1%. The 1918 Spanish Flu Pandemic has been estimated to have a CFR ranging from 2% to 20%, depending on which sources you check, but is most commonly given as 2.5%. This particular virus, depending on which statistics you look at. Considering only the number of deaths per infected individual, the USA, Mexico, and worldwide have CFR's of 0.110%, 1.97%, and 0.78% respectively.

These CFR's are likely overestimated as well; it's reasonable to assume that the actual number of cases is higher - probably significantly higher - than the reported number of cases. A certain number of persons with mild symptoms, and those - such as myself - who simply like to avoid MD's (Unfortunately, I can't avoid other Ph.D.'s, I"m surrounded by them) will not seek medical attention, and will simply go unreported. These unreported cases will lower the CFR.

These results are probably not statistically significant as of yet. Flu seasons often occur in waves; in the US, there are often two waves of flu activity, with one wave occurring in late November and lasting into January, while another wave often occurs in January, peaking in and becoming most severe in mid-February, and finally tapering off in late March.You can see this for yourself here.

A realistic CFR cannot be calculated until there is more data, and the infection has started to subside. It could be too early to say, but thus far, it doesn't appear that this strain is particularly virulent, and the number of new cases (according to my graph) seems to be tapering off. There is speculation that this virus could mutate and become more virulent, and this certainly could occur. In my opinion, this is not very likely.

Mutations can and do occur in viruses - generally more so than in complex organisms - however, the in the evolution of viruses, in general, is for them to evolve to become less virulent, not more virulent over time. Secondly, this particular strain of flu virus is no more likely to mutate than is any other flu virus, including the seasonal flu. Thus far, the data seem to indicate that the dreaded H1N1 swine flu is no more virulent than, and no more likely to cause death than the typical seasonal flu that comes and goes each year.

The WHO has taken the claims of Australian researcher Adrian Gibbs seriously, and investigated his claims that the H1N1 Swine flu virus was perhaps man-made, and was mistakenly released into the environment. The WHO asked a group of virologists to evaluate Gibbs claims; the group reports that there is no evidence that the strain is anything but a naturally occurring virus (Source).

"This group of scientists feels that the hypothesis does not really stand up to scrutiny," Fukuda said in Geneva, where the WHO is based. "The evidence suggests that this is a naturally occurring virus and not a laboratory-derived virus."

I spent a good deal of time looking at sequence alignments and Phylogenetic trees over the past couple of days, and there are really only two unique things about this virus. Firstly, the virus came out of Mexico, and came during what is not historically considered to be 'flu season.' The fact of the matter is that tens of thousands of people die every year from the flu, and none of us ever bat an eyelash. It's common to see deaths from the flu during flu season... not so much when it's not flu season. The second thing the struck me as bizarre is that when I BLAST the nucleotide sequences, they show up as strains that have appeared in both North America and in Europe. I speculate that this is probably what led Gibb to conclude that the virus was likely man-made.

 I'm no virologist, but I certainly understand enough about viruses to know that they do recombine all the time. I further am aware that when two different strains of a the same virus infect the same host, that recombination can and does occur. In other words the fact that the virus is comprised of genes from two strains whose origins derive from other continents is entirely unremarkable - especially given the amount of travel and intercontinental trade that occurs. All it would really take is for a European-bred pig infected with European swine flu to be brought to North America, and mixed with hogs infected with the N. American strain.

Other elements of the sequence are equally unremarkable. For the most part there is remarkable homology between Viral sequences derived from distinct locations, there are occasional single nucleotide polymorphisms, but this would certainly be expected. We would expect this for a number of reasons: Firstly influenza's genome is composed of RNA; RNA genomes are inherently more unstable when compared to DNA genomes, and mutate much more readily. Secondly influenza genomes are relatively small - about 12kB. The small size of the genome combined with the rapid replication and mutation rates of viruses means the genome is likely to accumulate these SNP's. In looking at the sequences, there are not a large number of differences between these H1N1 strains and other H1N1 strains. Gibb commented on this rate of mutation, claiming that it seemed like there was 'accelerated evolution.' I can't actually comment of the speed at which flu genomes mutate - or at least I can only comment enough to say that the mutation rates didn't seem out of order to me. However, I will say that recombination events can make determining something like the rate of evolution extremely difficult and somewhat arbitrary.

It seems that I'm in agreement with the WHO virologists as well. I'm certainly willing to suspend judgement until I've seen Gibb's analysis, and what specific details lead him to believe this is a laboratory accident, and not simply a natural recombinatory event. Incidentally, I'm not sure that there is any way to detect a difference. If a virus has been cultured and subsequently recombines in an egg in a lab, or in a host organism in the wild, the process is the same. The enzymes, the raw materials, and the pathways used to recombine the viral genomes are the same, and recombination in eggs vs. host organisms shouldn't generate notable differences.

On a final note: the Google 2009 H1N1 Flu Outbreak Map reveals some interesting things about epidemiology in action. Note that the outbreak began in Mexico and quickly spread into the US. The responses between the two countries were almost polar opposites, with Mexico taking a hardline, conservative approach to dealing with epidemic, and implementing various restrictions and forced closures. The US on the other hand took a more laisez faire approach and did nothing. As a result of these two policy differences, the number of cases in the US has surpassed the number of cases in Mexico (Source).

14 May 2009 -- As of 06:00 GMT, 14 May 2009, 33 countries have officially reported 6497 cases of influenza A(H1N1) infection. Mexico has reported 2446 laboratory confirmed human cases of infection, including 60 deaths. The United States has reported 3352 laboratory confirmed human cases, including three deaths. Canada has reported 389 laboratory confirmed human cases, including one death. Costa Rica has reported eight laboratory confirmed human cases, including one death.

In an article published on line today, it is reported that the World Health Organization (WHO) is investigating a claim that the swine flu virus may have resulted from human error.

Adrian Gibbs... intends to publish a report suggesting the new strain may have accidentally evolved in eggs scientists use to grow viruses and drugmakers use to make vaccines. Gibbs said he came to his conclusion as part of an effort to trace the virus's origins by analyzing its genetic blueprint. "One of the simplest explanations is that it's a laboratory escape," Gibbs said in an interview with Bloomberg Television today. "But there are lots of others."

There certainly are lots of others, the first thing that popped into my mind was the idea that this was an engineered virus. After all there are several peculiarities; for example, I've analyzed the DNA sequences myself, and they are from strains found in both North America and Europe. The article furthers this point:

Gibbs said his research found the rate of genetic mutation in the new virus was about three times faster than that of the most closely related viruses found in pigs, suggesting it evolved outside of swine.

So far, so good, right? Mmmmm.... I'm not so sure, if you keep reading in the article you eventually read the following:

Gibbs said he has no evidence that the swine-derived virus was a deliberate, man-made product. "I don't think it could be a malignant thing," he said. "It's much more likely that some random thing has put these two viruses together."

I've italicized the important text in the above quote: "he has no evidence" and doesn't "think it could be a malignant thing" Why the hell not? Something that can be done unintentionally in the laboratory can certainly be accomplished with intention. What evidence does he have that the virus isn't deliberately man made? How could he tell the difference? If the recombination were simply performed by mixing viruses together in eggs and isolating out particular strains there is not really any way that you could tell the difference.

An important question about this: If one chooses to speculate that the virus came from passage in eggs by vaccine development and viral production - and this point is disputed by other researchers in this article - where specifically in Mexico are these activities occuring? Is Mexico a hotbed of vaccine production? To be honest with you, I have no clue. However, if you're going to propose something such as this, you should look at the epidemiological data also. You should determine, as near as is possible, where the outbreak initiated, and whether or not there are any plausible places that such a passage may have occured. If you can't find such an institution, then - in my opinion - this increases the probability that this is in fact a deliberate engineering and release of virus into a population; in other words a bioterrorist event. In any case, I'll continue to monitor the situation, and perhaps perform some sequence analysis of my own.