Medical Examiner

Swine Flu and the Mexico Mystery   

Why does the swine flu seem to be more deadly in Mexico?

Click here to read more from Slate on the swine flu. 

Two weeks ago, no one had heard of this strain of swine flu. Now it’s on every front page and almost every continent. Is this the deadly global pandemic we’ve been worrying about?

Certainly there’s cause for sober concern, if not alarm. For of the two qualities vital to a nasty pandemic—to spread readily and to be deadly—this flu, a brand-new strain of swine flu, or H1N1, seems to possess the first: Evidence is strong that it spreads readily among humans. In that sense, it’s an inversion of the bird flu. Bird flu terrifies infectious disease experts because it kills about half the humans who get it—but it has so far failed to develop the ability to jump easily from person to person.

This swine flu, meanwhile, does seem to spread easily by airborne transmission. But how deadly is it? Despite the 100-plus deaths in Mexico, we don’t really know. And that’s why epidemiologists are working frantically to figure out the Mexico mystery: Why do the death rates there appear to be so much higher than those in the United States? In Mexico, it has reportedly killed about 100 of the 1,600 official suspected cases; elsewhere, it has appeared to take a far milder course, with zero deaths out of approximately 300 instances. There are several possible explanations for this discrepancy—any one, two, all, or none of these ideas could shed light on how deadly this virus might prove. In order of ascending likelihood for Mexico’s higher mortality:

1) Perhaps population-level genetic differences render the U.S. population more resistant to this strain’s effects than the Mexican population.

This suggestion has popped up on a few blogs. Does the more indigenous genetic makeup of many Mexicans make them more vulnerable? Though it’s nice to see people think genetically, the genetic differences in question would have to be far wider than they are to explain the differences. This isn’t like the smallpox situation of 500 years ago, when American Indians were decimated by a virus they’d never encountered while Europeans carried it easily because centuries of exposure had selected them for resistance. This strain of swine flu virus is apparently new to everyone—a combination of bird flu, seasonal human flu, and (predominantly) two kinds of swine flu, all in a form our bodies have never seen. There seems no reason any human population should resist its effects substantially better or worse than any other. We can probably put this “genetic vulnerability” explanation in a drawer.

2) We’re really looking at two different viruses, but WHO and the CDC haven’t picked up on it. 

This was a halfway plausible explanation before the full flu assays were done on significant numbers of both U.S. and Mexico cases. But the completed assays appear to show that the fatal Mexico cases match closely the 40-plus milder cases confirmed in the United States. It’s possible, but highly unlikely, that a high proportion of the other 80-plus suspected swine flu deaths in Mexico will prove negative. At least for now, we can probably set this possibility aside as highly remote.

3) Some secondary health issue present in Mexico but not elsewhere—another bug common in the population or in hospitals—is combining with the swine flu to make it more deadly there. 

This is remains a distinct possibility. The CDC has tested samples from the fatal cases in Mexico for some possible secondary bugs and vulnerabilities and eliminated the most likely and the easiest to test for, but there’s no blanket test for all such candidates. So it’s still possible some other bug joined with the swine flu to claim most of the fatalities. It’s also possible that Mexico City’s air pollution sharpened the course of cases there.

4) Some difference in the way we’re tracking and counting cases—a “surveillance difference”—is making the Mexico situation seem worse than it is and the U.S. situation seem better than it really is.

This is a virtual certainty—but with implications that are highly uncertain.

We can be sure we’re counting things differently. We don’t know the real numbers in Mexico, and the total caseload elsewhere is measured in dozens, which is small enough that you would expect only very high kill rates to show.

Most of the dissonance appears to be from Mexico, where the numbers are bigger and reporting apparently more chaotic. There are some suggestions that a combination of disorganization and politics has slowed or outright discouraged rapid identification and tracking of these cases once it was apparent something was afoot. Doctors in some hospitals are reporting more deaths than now noted, coverups, sloppy containment, and generally dire situations; others claim things are a bit more orderly.

And because Mexico had the bad luck to have its cases emerge in the midst of its flu season, the virus had generated a significant case load and quite a few serious cases before officials realized something more serious was occurring and started tracking in earnest.

All this affects the apparent significance of the numbers involved. Of the 110 million people in Mexico, 1,600 cases have been reported, with about 100 deaths—suggesting a mortality rate of 6 percent. This is almost certainly bad math, as the total case count almost certainly ignores thousands or tens of thousands of other cases that have taken milder courses like those in the United States. It’s perfectly conceivable Mexico has actually had 10,000 or 100,000 cases—or even 1 million cases. If so, then the kill rate would be not 6 percent but 1 percent (given 10,000 cases) or 0.1 percent (given 100,000 cases). * If it’s 1 million cases (quite possible if this thing really spreads easily) then the mortality rate is just 1 in 10,000. Meanwhile, because the United States is on high alert—and can take special note of people with recent travel to Mexico—it is probably picking up a fairly high percentage of its cases, including milder instances that would have gone unnoticed in Mexico a few weeks ago.

If it hasn’t infected that many thousands of people in Mexico, on the other hand, that would suggest that, though it may be deadly, it doesn’t spread as readily as we fear. To hear of multiple tourist groups coming down with the virus suggests it spreads like wildfire. But it also ignores the virtual certainty that many tourists and other travelers have been exposed without getting ill.

That’s not to be too sanguine. For one thing, it’s also possible that Mexico is missing, undercounting, or badly underreporting deaths. But if this virus really does spread rapidly, its kill rate is fairly low; and if its kill rate is anywhere near as high as the 100-out-of-1,600 suggests, then it doesn’t spread very easily.

AP video: Obama addresses the swine flu outbreak.

The answer doubtless lies somewhere in between. The CDC team that went to Mexico on Friday hopes to gather better numbers; with luck, they will soon get a decent grip on how many have likely been exposed, how many of those were infected, and whether a secondary infection or other factor might account for the deaths. That information will help us decide whether Mexico can be viewed as a reliable predictor of what will happen elsewhere.

As Obama’s team wisely stressed in its press conference on Sunday, pandemics, like the viruses that cause them, behave in dynamic and unpredictable ways. And just as every flu pandemic differs from every other, a given outbreak, like this one, can display different dynamics on Day 100 than on Day 30, which is roughly where we are now. The transmission dynamics change as more people in new places get it. And as the virus encounters new populations, new environments, and new bits of other flu bugs, it, too, willchange—for such is a virus’s nature—possibly becoming deadlier or more transmissible, possibly becoming less so.

As one blogger put it, “If you’ve seen one flu pandemic, you’ve seen one flu pandemic.” But figuring out exactly what’s happening in Mexico will give us a much better idea of what we can expect around the world.

Correction, April 29, 2009: This article originally miscalculated the potential kill rates of swine flu. (Return  to the corrected sentence.)