MONSTERS AT THE DOOR
IN HIS 1947 EXISTENTIALIST NOVEL THE PLAGUE, FRENCH NOVELIST Albert Camus paints a gripping picture of medical workers joining together to fight an outbreak of bubonic plague. Set in the Algerian port city of Oran, the characters in the story range across a broad swath of everyday life from doctors to fugitives to clergymen, all forced to address the very “Camusian” issue of the human condition and the vagaries of fate. The end result of these separate deliberations seems to be that humans have at best an illusion of control of their destiny, and that ultimately irrationality governs events. The Plague is an account of an event that lies so far outside the reasonable expectations of normal experience—of how life should be—that we see it as simply…well absurd, as Camus famously described the conflict between what humans nostalgically expect of existence and the realities of our quixotic, unpredictable, unbelievable world. In other words, Camus’s plague was an X-event.
The Plague is but one of a huge number of fictional accounts of an epidemic and its effect on the daily lives of a large population. The surface story line of Camus’s tale is that thousands of rats begin to die unnoticed by the residents of the city. Soon, a local newspaper reports this seemingly strange phenomenon and a mass hysteria develops in the populace. In a well-meaning, but tragic, effort to quell the hysteria, public officials collect all the dead rats together and burn them, thus creating a catalyst that actually promotes spreading of the plague. Following a lot of political bickering over what actions to take, the town is quarantined, postal service is suspended, and even phone and telegram service is confined to essential messages. How this latter edict contributes to confining the disease is left a mystery, but it certainly contributes to the sense of isolation felt by the townspeople.
As the plague spreads throughout the city, people eventually give up their petty individual concerns and join together to help one another survive the pestilence. Finally, the plague burns itself out and life returns to normal. People take up their previous daily life patterns, and gradually as life becomes routine, the sense of “the absurd” that the plague revealed is paved over. And so it goes.
In Camus’s day, it was relatively easy to confine a disease to a local geographic region, as people didn’t fly halfway around the globe for a long weekend in the Seychelles or buy food in their local market that began the day on another continent. But in today’s world the plague outlined by Camus would almost surely not be confined to the borders of Oran, but quickly spread to continental Europe and from there to Asia and/or North America and/or South Africa and/or…My task in this chapter is to look at the possibility for just such an outbreak and the likelihood of its decimating hundreds of millions of people (or more) before it’s run its course.
This story of how the plague spread in Algeria provides a lesson in complexity through the way the individual pieces of the story—the actions taken by the various parts of the city administration and the populace—combine to produce entirely unwanted and unplanned “emergent” effects like burning of the rats that actually contributes to spreading of the plague instead of containing it. So the real story of this disease that makes it a complexity-generated X-event is not the outbreak of the disease itself, but the way the human systems interacted so as to exacerbate the death toll rather than reduce it.
Before setting sail on this voyage through the world of viruses, bacteria, and other nasty, dangerous, and infectious things, let me first clarify a bit of terminology that I’ll use throughout this chapter.
With these definitions at hand, we see that epidemics and even pandemics are far from a new phenomena. They have been with us for just about as long as humankind has walked the planet. And they’re not going away anytime soon. Just to put some meat onto this skeletal statement, here is a short list of some of the more infamous and deadly outbreaks of such diseases over the last couple of millennia.
This chronicle could be greatly extended but the point is clear. Epidemics and their much nastier relatives, pandemics, richly deserve their position as one of the Four Horsemen of the Apocalypse. But the foregoing list is just a summary.
One might wonder where these killer diseases come from and whether they have been around since living organisms crawled out of the primeval soup. According to recent work by Nathan Wolfe, Claire Dunavan, and Jared Diamond, major human diseases are of fairly recent origin. For the most part, they have arisen only after the development of agriculture. This work identifies several different stages through which a pathogen that originally infects only animals can evolve into one that exclusively infects humans. The main point in this research for us is that diseases leading to epidemics can arise from sources that originally have nothing to do with humans, at all.
To understand the likelihood of another killer plague, we need more information about not just how these infections start, but also about how they spread through a population. To this end, let’s look at how a modern plague, Ebola fever, has unfolded over the course of the last quarter century or so.
SAME STORY, NEW CAST
IN 1976, MABAKO LOKELA WAS A FORTY-FOUR-YEAR-OLD SCHOOLTEACHER in Zaire. Returning from a trip to the north of the country in late summer that year, he became sick with a very high fever. During the next week, he started vomiting and began to bleed from his nose, mouth, and anus. He died less than a week later. At the time no one could pinpoint the cause of his death, although he is seen today as the first victim of what we now call Ebola fever.
Not long after Lokela’s death, more than three hundred other patients began turning up with the same symptoms. The overwhelming majority of them died within a couple of weeks. Thus, Ebola fever came onto the radar of the international medical community as perhaps the most virulent disease ever to infect humans.
Thirty years after the first outbreak, the precise origin of Ebola is still unclear, although some evidence points to fruit bats as the carrier. What is known is that the disease migrated somehow from the African jungle to the outskirts of Washington, D.C., in 1989, and a secret military SWAT team of soldiers and scientists was mobilized to stop the virus breaking out in the nation’s capital.
What does it take for a pathogen like Ebola to spread through a population? And what are the warning signs we should be looking for as a signal of an epidemic in the making?
The first point to note is that when it comes to infectious diseases, not everyone is created equal. Some people are simply better positioned genetically and socially to transmit the disease than others, their immune systems having the ability to tolerate the disease in its infectious stage long enough to pass it on before either succumbing to or recovering from the infection. In severe acute respiratory syndrome (SARS), a Chinese physician spread the infection to a number of people in a hotel, who in turn took the outbreak to other Asian countries. The disease ultimately spread to more than thirty countries around the world and killed over eight hundred people.
Epidemics are a function of the disease pathogen itself (the virus or bacteria), the people who actually have the disease, and the connective structure of the overall population in which the infected people circulate (the interaction patterns of infected and uninfected people). This process has a striking parallel to the spread of information throughout a population, in which an idea spreads from one person’s brain to that of another person instead of it being a virus or bacteria moving from body to body. Formally, the two processes are identical, other than that in one case the infectious agent may be a few bars of a popular song or a computer virus, while in the other it is an infectious biological agent.
Best-selling writer Malcolm Gladwell has described the process of the outbreak of an information epidemic in his book The Tipping Point, where he identifies three laws of epidemics: the Law of the Few, the Stickiness Factor, and the Power of Context. These so-called laws parallel similar principles used by epidemiologists to characterize and model the spread of a disease through a population. So let me briefly summarize each of them:
The Law of the Few: There exist “exceptional” people in a population who are extremely well connected and at the same time strongly virulent. As a result, these few special people have the ability to expose a disproportionately large number of the population to the infectious agent. In the lingo of the epidemiological community, such people are termed “superspreaders.” An outbreak of SARS in Toronto, for instance, was traced to such a superspreader.
The Stickiness Factor: This law says that relatively simple changes can be made to many pathogens enabling them to “hang on” in a population, year after year. Influenza is a good example, where each autumn new strains of last year’s virus appear, each a slight modification of what came before, the changes being just enough to enable the virus to slip through the immune systems of many people and infect a large fraction of the population.
The Power of Context: This law asserts that humans are a lot more sensitive to their environment than it may seem at first glance. In other words, whether people are ready to change their behavior and, for instance, voluntarily quarantine themselves or even take basic precautionary measures to avoid infection, such as wearing a mask or washing their hands, depends on the cultural standards of the particular population they are a member of. In a small town, people will react differently than in a major metropolis. And that difference may be the difference that matters insofar as whether an epidemic breaks out or not. This is a good point to interject a few words about where our ideas of complexity enter the pandemic story.
At the level of a growing complexity gap leading to an X-event, the picture is rather clear, at least for an individual. We have two systems in interaction, the pathogen and the human immune system. Each has its own complexity level, determined in one case by the tools that the pathogen can employ to penetrate the immune system’s defenses, pitted against the tools the immune system can bring to bear to resist the attack. As long as these two levels of complexity stay more or less in balance, no infection takes place. But when the pathogen mutates faster than the immune system can react, that’s when trouble begins. And as that gap between the two systems widens throughout a major fraction of a population, an explosive level of infection can occur. Eventually the gap is narrowed as the immune systems of the population finally adapt to the pathogen. But the speed of complexity increase on the two sides of this “arms race” may be very different, accounting for the many years it often takes for a plague-style pandemic to run its course. This arms-race type of complexity gap is at the level of individuals. But there is also a population-level complexity story, as well.
The three network principles outlined above by which infectives interact with those who are not infected and pass a virus or bacteria to them is a network complexity issue. In particular, studies by network analysts like Duncan Watts and Albert-László Barabasi have shown that there are critical levels of connectivity in the linkages among the population at which an infection can suddenly “take off” like a wildfire. The threshold between containment of the disease and its going “viral” (that is, literally out of control) is a very fine line, an illustration of the butterfly effect I discussed in Part I.
So these are the rules of the game by which an epidemic of either a disease or a rumor breaks out and spreads. What are the stages we should be aware of that give us an early-warning sign of an epidemic in the making?
According to the World Health Organization (WHO), an influenza pandemic has six distinct phases, ranging from the appearance of an influenza virus subtype in animals with low risk of human infection in Phase 1, to sustained transmission of the virus in the general human population in Phase 6. The various phases constitute an increasingly clear set of signals, or “fingerprints,” that a pandemic is brewing. Here is a summary of all six phases:
Phase 1: No new influenza virus subtypes have been detected, but an influenza virus subtype that has caused human infection may be present in animals. If present only in animals, the risk of human infection or disease is considered to be low.
Phase 2: No new influenza virus subtypes have been detected in humans. However, a circulating animal influenza virus subtype poses a substantial risk of human disease.
Phase 3: Human infection(s) with a new subtype, but no human-to-human spread, or at most rare instances of spread in a close contact.
Phase 4: Small cluster(s) with limited human-to-human transmission but the spread is highly localized, suggesting that the virus is not well adapted to humans.
Phase 5: Larger cluster(s) but human-to-human spread is still localized, suggesting that the virus is becoming increasingly better adapted to humans but may not yet be fully transmissible (substantial pandemic risk).
Phase 6: Increased and sustained transmission in the general population.
As an example of the use of this list to characterize the stage of a possible pandemic, the so-called bird or avian flu, technically labeled the H5N1 virus, is currently at Phase 3. A movement up to Phase 4 would represent a huge increase in the danger to humans, as it is the first phase at which human-to-human transmission would be confirmed. This would no longer leave the virus in the careful monitoring stage, but enormously elevate the importance of seeking a vaccine and initiation of preventive public health measures.
PUBLIC HEALTH, PRIVATE LIVES
AS RECENT LEGISLATION HAS SHOWN, HEALTH IS NO LONGER A PRIVATE matter. For instance, due to the health hazards of inhaling even secondhand cigarette smoke, the United States and many nations in the European Union (but alas, not Austria) have banned smoking in all public places, including restaurants, bars, and cafés, on the grounds of health. Bear in mind here that secondhand cigarette smoke doesn’t have nearly the infectious or immediate life-threatening potential of something like Ebola fever, Spanish flu, or even tuberculosis. So where do we draw the line between the limitation of personal freedoms and public health?
A particularly graphic illustration of this dilemma occurred in the early 1900s with the cook Mary Mallon, known to history as “Typhoid Mary.” She was an immigrant from Ireland who worked in the New York City area between 1900 and 1907. During this period, she infected more than two dozen people with typhoid fever, even though she showed no signs of the disease herself.
People catch typhoid fever after eating food or drinking water that has been contaminated through handling by a human who is a carrier of the disease. Mary Mallon almost surely had had a bout of typhoid herself at some point in her life, but the bacteria survived in her system without causing further symptoms.
When public health authorities confronted Mary with the news that she was a possible carrier of the disease, she strongly denied several requests for urine and stool samples. Part of her argument was that a local chemist had tested her and found she had no signs of the disease-causing bacteria, at least at the time of the testing. Eventually, the New York City Health Department placed her in quarantine, isolating her in a hospital on North Brother Island for three years. She was released under the condition that she would no longer work in the preparation and serving of food.
But Mary was having none of this; she adopted the pseudonym “Mary Brown” and returned to work as a cook. In 1915, she infected twenty-five people at Sloan Hospital in New York, after which she was again taken into custody by the health authorities and returned to quarantine where she spent the remainder of her life. Typhoid Mary died in 1938—of pneumonia, not typhoid—and was cremated.
The case of Typhoid Mary illustrates perfectly the ethical dilemma facing public health officials: How do they “properly” balance the rights of Mary Mallon to freedom of movement and employment with the rights of the public to be protected against life-threatening actions and behaviors by other people in society? This is the intrasociety dilemma. There is an extrasociety version, as well: How does a country balance the right of movement of people across its borders with the right to protect their nation from emerging infections? Let’s look a bit more at these two situations.
In late 2006, the WHO announced an outbreak of tuberculosis (TB) in the KwaZulu-Natal region of South Africa. Alarmingly, of the 544 patients in the WHO study, nearly 10 percent had a new strain of TB that was resistant to not only the so-called frontline drugs, but also to at least three of the six “backup” treatments. The median survival time for these extensively drug-resistant TB (XDR-TB) cases was just sixteen days.
Coupled with the high level of HIV infections in the country, South Africa also suffers from a huge level of infectives who fail to comply with the drug regimes they’re given to cure the TB. WHO estimates that 15 percent of patients fail to complete the frontline programs, and a staggering 30 percent default on backup drugs. This has led to an overall cure rate of only about half the patients, which has made XDR-TB not only a potential national disaster in South Africa, but threatens the world at large, via South Africa’s steadily increasing tourist population.
To stem the spread of XDR-TB, a number of very severe social measures have been proposed, ranging from restoring social welfare benefits to hospitalizing patients so as to encourage them to remain hospitalized to far more extreme measures such as forcibly detaining people with XDR-TB. The WHO currently recommends that such patients voluntarily stop mixing with the uninfected population. But there are no measures to enforce this separation. The South African government has thus far not been willing to employ detention as a public health measure. All this despite the fact that international law does allow for precisely this type of forcible restraint if all other measures to stop the spread of disease have failed.
So the situation in South Africa with respect to XDR-TB is a living example of the conflict between the removal of an individual’s freedom of motion and right of assembly so as to shield the general population from a killer disease.
Here is another huge threat looming on the horizon.
Chinese hospitals derive a substantial fraction of their income from selling drugs to patients. As a result, doctors routinely prescribe multiple doses of antibiotics for routine problems like sore throats. This has led to a dramatic increase in the evolution of antibiotic-resistant strains of bacteria.
Warnings are already being given about the spread of these new strains through international air travel and food distribution, as Chinese pigs imported into Hong Kong in 2009 already showed signs of being infected with these “superbugs.”
As super-resistant strains of bacteria start appearing at ports of entry around the world, nations are starting to face ethical conundrums when it comes to sealing their borders against travelers and immigrants possibly carrying a communicable disease. Is there anything a country can do to protect itself from this type of threat?
During the SARS outbreak, the government of Singapore installed thermal-imaging scanners at all ports of entry into the country—sea, land, and air. The body temperature of everyone entering the country was checked prior to their going through immigration to see if they had a fever. This was a simple, nonintrusive screening procedure, comparable as a nuisance factor to normal airport security checks. But we can’t say the same for other possible measures to control the importation of a disease at a national border.
In the United Kingdom, there have been calls for compulsory screening of all immigrants for TB and HIV. While debatable as an effective measure for keeping these diseases from crossing the border, there is no debate whatsoever on the practical and ethical questions such a procedure raises. For example, which immigrants are singled out for screening? All of them? Just those from certain countries? Only asylum seekers?
We see that such border “filters” give rise to the possibility of discrimination, loss of privacy, and a certain kind of stigma. To put it compactly, stopping diseases at the border is not quite the same thing as controlling immigration.
So what can we do realistically to prevent a pandemic?
There are at least three ways to stop a pandemic:
Eliminate Infected Animals: By slaughtering the entire poultry population of one and a half million birds, the authorities in Hong Kong stopped the avian flu virus H5N1 in its tracks after the initial cases of human infection were reported in 1997. Unfortunately, this process was both hugely expensive and not totally effective, since the virus has reappeared since then. Nevertheless, the procedure has some measurable effect, at least if the virus can be localized. Similar mass culling was used in the United Kingdom in 2001 to stem foot-and-mouth disease in cattle, when four million animals were slaughtered. But this “shotgun” approach to stopping disease raises many troubling questions, not the least of which is who is going to compensate the farmers for the loss of their animals, hence their livelihood?
Vaccination: Protecting animals and humans by vaccination is also a tricky business. For instance, even if a vaccine exists, it may be impractical to administer it to large numbers of people or animals. Moreover, it’s difficult to distinguish a vaccinated animal or human from one that’s not. So restriction of movement of the nonvaccinated may be difficult to monitor or control.
Drugs: In contrast to vaccines, which are preventive measures, drugs are an after-the-fact treatment to prevent the outbreak of a pandemic. For bacterial infections, there now exist many very effective antibiotics. There also exist a growing number of bacterial strains resistant to such drugs, like the XDR-TB noted earlier.
When it comes to viruses, the situation is far worse. The only effective antiviral agent for the avian flu virus H5N1 seems to be Tamiflu, which acts as both a kind of vaccine in that it prevents infection, as well as a drug that enhances the survival rate of those already infected. But in either case, it must be given very shortly after exposure to the virus or contraction of the infection. Moreover, variants of the virus have already turned up that are resistant to the normal Tamiflu treatment. So yet again there exist no magic bullets for all known infective agents.
Strangely, perhaps, the most effective general procedure for stopping an outbreak from becoming a full-fledged pandemic is simple common sense. The key element is educating the population about elementary procedures for health care and sanitation. For example, washing your hands when handling food, keeping your home and outdoor areas clean, properly taking medications, and other such procedures go a very long way toward stopping infectious diseases before they can develop into a pandemic or even into an epidemic.
But what about stopping pandemics before they have a chance to get off the ground? Do we have any procedures for effectively forecasting the outbreak of something like avian flu or SARS? This takes us into the realm of how to model the way an epidemic or pandemic takes place once an infection has gotten a foothold in a population. So let’s look at some unexpected directions that researchers are taking in understanding the way diseases spread both in space and in time.
PLAGUING PATTERNS
VIRTUAL PLAYGROUNDS, LIKE THE ENORMOUSLY POPULAR MULTIPLAYER online games World of Warcraft or Second Life, have a following that numbers in the hundreds of thousands. In World of Warcraft, players interact in real time on the Internet using computer-controlled avatars to fight battles, form alliances, and gain control of territory.
At first glance, World of Warcraft hardly seems like a testing ground for real-life battles against influenza, SARS, bubonic plague, or any other type of communicable disease. But first impressions can be deceiving. And work in the United States by Nina Fefferman of Rutgers University and her collaborator, Eric Lofgren of Tufts University, is showing how these virtual worlds can provide deep insight into the way pandemics form in the world we actually inhabit.
For many decades, mathematical epidemiologists have been creating models of the spread of disease, in order to try to understand and predict the outbreak and spread of epidemics. Unfortunately, to make these models mathematically tractable it’s necessary to introduce a host of simplifying assumptions that sometimes assume away the very questions they’re trying to answer. So computer games, which allow an almost unlimited variety of detailed behavior by the players to be incorporated into their actions, seem to be a good way to overcome some of the limitations of the math, say Fefferman and Lofgren.
The collaboration between the two scientists and the game’s publisher Blizzard came about when programmers introduced a highly contagious disease into a newly created zone in the game’s hugely complicated environment. Initially, the “patch,” as such new elements are termed, worked as envisioned: veteran gamers recovered from the disease, while inexperienced players succumbed and were left with severely disabled avatars.
But soon things began to go out of control. Just as we see in the real world, some of the infected avatars made their way into the heavily populated cities in the virtual world and infected the inhabitants. The disease also spread through infected domesticated animals, who were quickly abandoned by their owners and left to roam aimlessly through the world infecting other animals and avatars. In short, it was a virtual pandemic.
Programmers at Blizzard tried setting up quarantine zones. But in the virtual world, as in the real world, quarantines were ignored as avatars tried to escape so as to carry on their battles. Finally, the programmers had to shut down the servers and reboot the system in order to eliminate the disease and make the game playable again. Roll back the system! Wouldn’t it be nice to be able to do that in reality.
Lofgren was actually playing the World of Warcraft when the plague struck, and he immediately saw the potential for using the game as a testing ground for studying the spread of disease. What intrigued the researchers was the opportunity to study how people really behave in a public crisis as opposed to the behavioral assumptions made in the earlier mathematical models. People are very different from the homogeneous agents that populate the worlds of the mathematical epidemiologists. In those mathematical models, the individuals in the population all had the same properties regarding virulence of their infection, ability to infect others, and so forth. The heterogeneity possible in the computer models can make a huge difference as to whether a disease breaks out into an epidemic or not, argued the researchers. How many will try to escape a quarantine? How many will start to cooperate if they are scared, as in Camus’s story? As Fefferman says, “We simply don’t know.”
This is where the virtual worlds comes into play, since the players can be given individual characteristics for virulence, resistance to infection, cooperation, escape, and so on. The system can then be “turned on” to see what happens. Skeptics argue that players might be much more ready to take risks in the virtual world than in the real one. The counterargument is that players have invested considerable amounts of time and energy into making their avatars strong and in forming alliances. As a result, a lot of the players’ egos are invested in their virtual representative, and they don’t want to see their egos crushed by taking outrageous risks.
In the end, of course, the virtual world simulation is just that: a simulation. Like any model, it is not a perfect mirror of the real world. There are assumptions built into the virtual world, too. Still, it seems to be a promising step toward understanding how potential pandemics spread and, most important, how they can be stopped before they have a chance to get off the ground.
ADDING IT ALL UP
BEFORE SUMMARIZING WHAT WE’VE DISCOVERED ABOUT PANDEMICS, I’ll address very briefly a point that appears regularly in the popular press and elsewhere regarding pandemics: the question of bioterrorism.
Everyone can agree that bioterrorism is a potential problem. No doubt about that. And it’s a problem that perhaps deserves even more attention, or at least more resources, than it currently receives from governments around the world. But from the standpoint of my goals in this chapter, it doesn’t really matter much whether a pandemic arises from accidental or intentional human actions. The dynamics of the spread of disease and the end result are indistinguishable. For that reason, I have said nothing in this chapter about detection, prevention, and/or mitigation of terrorist attacks using biological weaponry. Now back to our story.
We have seen that even without the benefit of terrorists, nature is perfectly capable of brewing up a vast array of biological threats to human existence. Epidemics and pandemics of a bewildering variety have regularly made their appearance on the historical stage and can certainly be expected to reappear in various guises again. That goes almost without saying. The real question is whether humankind will be prepared to deal with a major pandemic when and if it occurs.
As to that question, the outbreak of a worldwide life-threatening disease can happen anytime. In fact, sooner is more likely than later for the twin reasons of the worldwide trend toward migration to the cities, thus leading to greater urban population densities, together with the rather lackluster international cooperation on the formation of monitoring and prevention of disease. People simply don’t want to take seriously another outbreak of Spanish flu or SARS or avian flu or…. But they are “out there.” And they will get you—if you don’t watch out!