Friday, October 28, 2005

READER BILL ARONSTEIN saw a previous post where I linked to the comments of one Instapundit reader, Patrick Cunningham, who claimed that the avian flu risks might be somewhat over-hyped. He's a medical researcher as well (PhD MD), and sends this long and interesting comment that I'm publishing with his permission:
In your recent post on the possibility of an avian flu pandemic, you mentioned a reassuring e-mail that Professor Glenn Reynolds published at his Instapundit weblog.

I think that several of the assertions made in that e-mail are overly optimistic, and we should not rely on them when thinking about how to plan for the possibility of a pandemic.

The first optimistic assertion is that: “The main way that flu kills is by predisposing its victims to "superinfection" by bacterial illnesses - in 1918, we had no antibiotics for these superimposed infections, but now we have plenty.”

Although it may be true that during ordinary annual epidemics of influenza, most of the patients who die do not succumb to viral pneumonia, but to superimposed bacterial pneumonia, that was decidedly not true in 1918-1919, and might not in general be true for episodes of pandemic flu. Many accounts of patients who became ill with the flu during the pandemic have been published. If you look at the course of illness in patients who died in the 1918-1919 epidemic you will note that many patients died within days of becoming febrile. Sometimes they died with 24 to 48 hours of first becoming sick. That course just doesn’t fit with the notion that they died of superimposed bacterial pneumonia after having been weakened by viral infection. They undoubtedly died from severe viral pneumonia and “ARDS,” the “acquired (or adult) respiratory distress syndrome” caused by their influenza viral pneumonia. Undoubtedly there were those who died with superimposed bacterial infections, but among the young and healthy, I think that flu itself was the leading cause of death. Why some people become sicker than others is not fully understood.

“Such superinfections, and the transmittal of flu itself, were aided tremendously by the crowded conditions and poor sanitation of the early 20th century - these are currently vastly improved as well.”

Compared with 1918-1919, the world is dramatically more urbanized, and urban areas dramatically more crowded today. In many parts of the world, sanitation is probably no better than in was then. Moreover, I think that this blithe confidence in the admittedly improved sanitary conditions in the United States overlooks the plain fact that the flu is spread in respiratory droplets and by person-to-person contact, not by contaminated water, nor by the absence of hygienic toilets. The improvement in overall sanitation is not relevant to influenza transmission. Indeed, we see flu epidemics every year that are limited only by the prevalence of at least partial immunity in vast sectors of the population. If modern hygiene were sufficient to prevent an avian flu pandemic, then we would not need to worry about annual flu epidemics.

“Flu hits the elderly the hardest, but the "elderly" today are healthier, stronger, and better nourished than ever before.”

It is true that during annual epidemics of influenza, most deaths occur among the very, very young and the very, very old. This creates the well-known “U-shaped” mortality curve when displayed on a graph. However, in 1918-1919, there was another huge spike of mortality among healthy patients in their ‘20s and ‘30s, thus producing what has been called a “W-shaped” mortality curve. Being healthy was no protection against influenza mortality during the pandemic, and might not be protective today. Scores of prominent, even world-famous, affluent individuals in the prime of life were cut down by influenza during the 1918-1919 epidemic, along with tens of thousands of other previously healthy people. Military recruits in large camps were particularly vulnerable.

“Our medical infrastructure is vastly better off, ranging from simple things like oxygen and sterile i.v. fluids, not readily available in 1918, to complex technologies such as respirators and dialysis.”

It is true that the present medical infrastructure is vastly different than the infrastructure that existed in 1918-1919, but the medical infrastructure is already strained. Most hospital ICUs operate at full capacity or very nearly at full capacity, and it is not economically feasible to maintain very much reserve capacity. Even the annual visitations of ordinary epidemic flu severely strain the hospitals and clinics in American cities that are hit with large numbers of infections during an active flu season. A pandemic, with vastly increased numbers of sick, would entirely overwhelm the system, and it is not realistic to rely on the assumption that respirators and dialysis machines would be available for the horrifying number of severely sick and dying otherwise patients in their 20s and 30s, let alone for the elderly or chronically ill.

“Should we be concerned? Sure, better safe than sorry, and concerns about publishing the sequence are worth discussing. Should we panic? No - my apologies to the fearmongers, but we will never see another 1918.”

Whether we see a worldwide pandemic on the scale of 1918-1919 depends on one thing only, the existence of a novel influenza virus which becomes easily transmissible. None of the mitigating factors upon which you so confidently rely will be sufficient to alter the overall picture, neither in the United States nor in the world.

The two factors that are important to estimate are the attack rate and the case fatality rate of avian flu infection. The attack rate means the number of people in a given population (town, city, country, etc.) who become infected. The case fatality rate means the number of infected people who die as a result of their infection. Thus far, it is assumed that all or almost all human cases were contracted from birds, and the attack rate has been low. Conversely, the case fatality rate has been extraordinarily high – something like 50%.

The case fatality rate is undoubtedly exaggerated because only the most severely affected individuals are being counted. There may very well be many more people who have more mild infections, who are not being tested, and not being counted.

At present, no official source is admitting the possibility of human-to-human transmission of H5N1 avian influenza. If that becomes possible, and the virus begins to spread, it is likely that the attack rate will be on the high side, since most people probably have no immunity to this particular combination of influenza virus antigens. I think that most people assume that most flu pandemics have an attack rate of about 25%. The attack rate for annual flu epidemics is estimated by the W.H.O. to be on the order of 5% to 15%.

The case fatality rate is, as I said, very difficult to estimate. The virulence of flu virus strains is highly variable. The case fatality rates in the 1957 and 1968 pandemics were not really any worse than usual for annual epidemic flu. The 1918-1919 virus seems to have been more virulent, and more lethal. Recently, scientists re-created the 1918-1919 virus by obtaining viral DNA from the corpses of patients who had been buried in permafrost during the epidemic. When inoculated in mice, the re-created virus proved much more deadly than usual flu viruses. This finding suggests that the case fatality rate for the 1918-1919 flu was truly extraordinary. I think that most scholars think the worldwide case fatality rate in 1918-1919 was about 2.5%, but it was much higher than that in certain areas. The case fatality rate for the usual annual flu epidemic is between 0.2% and 0.5% -- so the 1918-1919 flu was approximately ten times more lethal than the usual flu.

News is breaking daily, and I do not know if H5N1 is going to develop into a human flu pandemic. I think that it is certainly possible that it will do so. We should think about the possibilities realistically. I don’t think we should panic. But we also should not comfort ourselves with overly optimistic misperceptions.
He also emailed Dr. Cunningham, who he knows [see update -- FA] , in case he wanted to do any further comments. Will publish his reply if he sends one.

Of course, I'm not in a position to judge the merits of both argument, being a layman in this field. So I'll let the experts talk.

UPDATE. Bill Aronstein tells me that he actually doesn't know Dr. Cunningham; I inferred it from the fact that he was emailing him, but I was wrong.

Click here to send me an email