Effect Measure

Cholera, bird flu and humility in science

The “father” of epidemiology is a nineteenth century doctor, John Snow. He had more than one disciplinary child, since he is also considered the “father” of anesthesiology, having popularized the use of chloroform in obstetrics by using it on Queen Victoria in the 1850s. That distinction aside, Snow is famous for his pioneering studies showing cholera was a waterborne disease. I’ve been thinking about this in relation to bird flu. Here’s the connection.

In Snow’s Victorian London, the predominant scientific theory on cholera’s etiology was from miasmas, a general term for noxious elements in the atmosphere, usually produced by decaying organic matter and waste. The miasma theory as explanation had a lot going for it. It explained, for example, why disease epidemics, especially cholera and typhoid, seemed to concentrate in the poorest parts of town, where sanitation was worst. In a city like London, drowning in its own excrement, the smells and conditions of the poorest parts of town don’t even seem credible today, exceeding the worst barrios anywhere. There was no domestic sewer system to carry away the wastes of millions of residents resulting in overflowing cesspools, cellars and eventually the river Thames, running through the middle of the city. The miasma theory connected smells with disease, and bad smells were the rule, not the exception, in Victorian London. It was not surprising the city would be the locus of devastating cholera epidemics in the 19th century. While there were a variety of theories as to how foul air could cause disease (general debilitating effects, specific gases, electrical influences, etc.), the principal view of the mode of transmission was through the air, not the water.

Steven Johnson has a new book (The Ghost Map) on Snow and his effect on how we view the city. It is a pleasant read (if you find reading about fecal contamination and sudden death pleasant, in which case maybe you should consider epidemiology as a career), although not revelatory for those who know their public health history. But I found this observation about the miasma theory especially poignant:

So often what is lacking in many of these explanations and prescriptions is some measure of humility, some sense that the theory being put forward is still unproven. It’s not just that the authorities of the day were wrong about miasma; it’s the tenacious, unquestioning way thy went about being wrong. An investigator looking for holes in the theory could find them everywhere, in even in the writings of the miasmatists themselves. (Steven Johnson, The Ghost Map, p. 125)

Snow and others saw the same thing, but interpreted it differently. The scavengers, sewer hunters and “night soil men,” those who made a living combing the city’s waste to eke out a living on the margins, were not unusually sick. On the contrary, they appeared to be unusually robust. More significantly, only some of the people who lived under the exact same atmospheric noxiousness came down with the disease. A vast public health campaign to eliminate cesspools (and thus divert the waste into the river) had not solved the problem, which continued as before or worse. In defense of the miasmatists, the true source of the disease was essentially invisible while the truly noxious sanitary conditions of the hardest hit population was all too sensible and visible. The non-uniform pattern of attack had a ready explanation: individual susceptibility. This could be framed in many ways, depending on one’s social outlook. It might have reflected a moral failing or secret sinfulness of the victim, or a moral failing in a society that allowed poor people to live in such squalor. We have our modern versions of this and we are helped by handwaving but plausible arguments from immunology and genetics.

The techniques that Snow used to such good effect, statistical correlations between mortality and risk factors, also seemed to confirm the truth of the miasma theory. Deaths were greater at the lower elevations (which were also closer to the River and contaminated water sources). The same day that Snow convinced local officials to shut down a contaminated well on Broad Street that had led to an explosive outbreak around Golden Square, the General Board of Health issued detailed instructions for an investigation of the cause of the outbreak which involved collecting information, almost all of which was germane only to the miasma theory. The idea of compiling and analyzing statistical data was the same as Snow’s. But the questions asked dictated the possible outcomes.

What’s the connection with bird flu? If you use your word processor to do a Global Search and Replace of “almost all human cases of bird flu so far have been from close contact with poultry” for “miasma theory” you’d have somethingsimilar to what we read today. I have said here before that I still consider a poultry source a reasonable epidemiological position, since the causative agent, influenza A/H5N1 is widespread in poultry and exposure is prevalent. But humility is also called for. There is an abundance of human cases for which a good history of poultry exposure is difficult to ascertain and establish. On the other hand, exposure is so widespread, many wonder why there isn’t more disease in identically exposed individuals. The investigations naturally (and not unreasonably) ask questions pitched strongly toward poultry exposures. As in Victorian London’s unsanitary conditions, poultry exposures are common in the areas of the world where this virus is endemic, so investigations naturally reveal a connection (the example of the General Board of Health cholera investigation in 1855 comes to mind). Again the idea of individual susceptibility is invoked as the explanation, although in a much more sophisticated form buttressed by modern genetics and immunology, but with no more hard evidence. Generalized explanations about moral weakness had special cogency for the Victorian sensibility. Any explanation that uses the language of genetics, molecular biology and immunology has special cogency for us. We believe there are good and rational reasons for this cogency, but then the Victorians also believed the same about their spontaneous and transparently obvious (to them) moral reasoning. Again, it is a question of a little humility.

If it’s not poultry, then what is it? This isn’t an either/or question. It can be poultry and other things. The things that still seem to be out there, albeit at the margins for most epidemiologists who work on flu, is gastrointestinal exposure (food and water) and another vector or natural reservoir besides birds. Another infectious risk factor might also be involved, i.e., the risk of human illness is greater when there is exposure in combination with another agent or pre-existing infection. There’s not a lot of evidence for any of this at the moment. There is also some counter evidence for the poultry hypothesis, evidence that is convienetly ignored. Someday maybe it will be explained.

But the example of John Snow and waterborne cholera should remind us humility is also a scientific virtue.

Comments

  1. #1 Susan Och
    November 29, 2006

    Thank you. I have been thinking along these same lines but haven’t been able to articulate these thoughts nearly as well.

    Perhaps good science moves on a different timeline than sound-bite journalism.

  2. #2 Joe in Australia
    November 29, 2006

    Revere: I know you said this to deprecate it, but I want to add my criticism: “almost all human cases of bird flu so far have been from close contact with poultry”

    I only read the popular reports, but I haven’t seen many cases where the infection is actually tied back to an infected bird. For all I can tell the sentence should be “almost all human cases of bird flu so far have had close contact with the ubiquitous poultry of villages”.

    I agree that the poultry explanation is the best one we have at this stage but the evidence for it is … paltry. I feel that our infection profile is like the Australian bush in mid summer. The conditions are right for a bushfire, and there are an infinite number of places where a fire may break out. Tiny almost-fires light and extinguish in a thousand places, lit by careless hikers or lightning strikes or the sun’s rays focussed by some broken glass. Sooner or later one will spread successfully and we will ask why it began there rather than somewhere else. The answer is that it nearly began everywhere.

    I think it’s the same with the bird flu. There may be almost-right strains everywhere, perhaps even in asymptomatic human carriers. Sure, poultry are a viral reservoir but how do we know what fraction of the reservoir they represent? Sooner or later one of the almost-successful viruses will hit the sweet spot of mutation and circumstance that lets it cause an epidemic, and we’ll be asking ourselves why it happened there. I think it’s very important that we don’t supply an automatic answer.

    Disclosure: I had chicken for dinner last night.

  3. #3 Tom DVM
    November 29, 2006

    Joe in Australia. Every once in a while, I read a post that is exceptional.

    The post above is a beautiful piece of work…

    …clear…concise…accurate…

    Thanks.

  4. #4 serge
    November 30, 2006

    Thank you for your post but you write ::

    ”There is also some counter evidence for the poultry hypothesis, evidence that is convienetly ignored.”

    Convienetly ignored … maybe you have you could show us the evidence and how it is ignored.

    Thanks

  5. #5 revere
    November 30, 2006

    serge: It is mentioned in this and numerous other posts. A sizeable fraction of the cases have no history of close contact with infected poultry and a huge number of people who do have contact don’t become ill. I’ll invoke you as an example of how it is being ignored.

  6. #6 revere
    November 30, 2006

    Joe: In essence you are advancing a probability argument. I have no problem with it. It is a perfectly reasonable and plausible explanation. But it isn’t the only one and it has no more evidence going for it than most of the others.

  7. #7 jspreen
    December 1, 2006

    What’s the connection with bird flu?

    There is no connection. Read a book about early industrial revolution and the extremely difficult conditions of the poor and anybody with a normal brain can understand that decaying organic matter and waste were certainly not the most difficult problems the poor had to handle.
    19th century London cholera has nothing to do with 21st century avian flu. The avian flu pandemic is a stupid hoax invented by WHO and other officials who simply swap labels here and there to make their cases look the way they want them to look.

    What? Hundred of millions of dead birds? Yeah, killed by men. One chicken infected, a hundred thousand slaughtered.

    JS

  8. #8 Allan
    December 1, 2006

    Re. convenient ignoring of contrary evidence, is it not still the case that H5N1 genetic sequences from human cases in Indonesia are not the sequences found in poultry there? If this is still the case, it seems to me that at least WHO is conveniently ingnoring genetic evidence of something other than contact with poultry being the source of many human H5N1 infections in Indonesia.

    Keep up the good work, this is an important issue.

  9. #9 Margot
    December 2, 2006

    Revere, Your post on Cholera, miasma, etc is a truly thoughtful one and reminds me to keep in mind the vast uncertainties within bird flu epidemiology whenever I am tempted to believe that someone somewhere has found “the” key. I am not an epi-trained scientist, but have been reading the vast literature on H5N1 and its cousins for some time. YOur point about the “close contact” issue is well taken. However, what if asking about “close contact” is the wrong question? We have no idea how the histories are taken when patients turn up at clinics in Indonesia or Thailand or China, or anywhere else for that matter. It is easy to imagine that nearly everyone in Asia, at one time or another, has walked through a live poultry market. As we know, Viral particles from poultry are embedded in fecal matter which becomes dry and dusty and can be aerosolized under certain conditions. Poultry factory workers carry this stuff out every day. So asking about “close contact with poultry” is an incomplete question. Dusty shoes may not be sufficient to cause disease, but could cause low grade infections that are never reported. IN some people, though, particularly children, isn’t it possible that fecal dust may be enough for lethal exposure without immediate and close contact with the poultry?

    Again, however, we have the puzzling question: why are there not more cases of human infection? The answer may be that there are many more, but they are asymptomatic or so mild that they never come into clinics or hospitals. Also, it is possible that just as some of the poultry workers who were tested for H5N1 antibodies and found to be seropositive had never become sick, millions of people may have gradually acquired some immunity to variations of LPAI and even HPAI because of their long term exposure to birds, poultry, ducks, geese, etc over their lifespan. Mortality from H5N1, so far, has clustered among children and young teens between the ages of 9 and 16. IF they have not had much casual contaqct with infected flocks or live birds in market cages, one whopping exposure to H5N1 might be enough to overwhelm them. I think your points about caution and humility are absolutely on the mark. I also think it is critical that we apply more specificity to the questions and seek more detailed histories, and also conduct the background seroprevalence studies that the WHO now seems, finally, to be calling for.

    I enjoy Effect Measure’s contributions ever so much. Thank you.

  10. #10 revere
    December 2, 2006

    Margo: WHO has only just now advised on how to work up an outbreak. It has been a persistent problem that there is no standardized way to collect information (clinical, epidemiological, exposure, whatever) and then harvest it to see what is going on. That is a function of the places and conditions under which these infections are occurring as much as anything. Even in the US we don’t have a good medical information system for this kind of purpose. The alleged histories are often someone from the Ministry of Health saying “there were dead birds in the area.”

    About asymptomatic or inapparent infection. I think most of us thought that for a long while and I still think it is a possibility. But the seroprevalence surveys so far have failed to turn up any significant evidence of it. It is really quite shocking given the case fatality ratio. So I no longer know what to think.

  11. #11 Not My Second Opinion
    December 5, 2006

    Unfortunately, the connection you make between John Snow’s cholera and our pending Avian Bird Flu pandemic doesn’t go far in fixing our current problem.

    We cannot steal a water pump handle to halt H5N1. :)

Current ye@r *