David Brooks has written a New York Times column that is very misleading at best and dangerously irresponsible at worst. I know a little bit about the economics of the issue, and wanted to comment. If anyone in healthcare, epidemiology, or biology wants to add more commentary, that is very welcome.

Globalism Goes Viral
By DAVID BROOKS
Published: April 27, 2009
http://www.nytimes.com/2009/04/28/opinion/28brooks.html?ref=opinion

Brooks appears to use the current international outbreak of a new strain of influenza to argue against a robust international organizational effort to monitor and control newly emerging infectious disease.

The column might be an effort to take the edge off of the embarrassment of the GOP for attacking, a few months ago, funding of efforts to control pandemic flu in the stimulus package (I’m guessing… it is difficult for me to understand what Brooks is trying to do). But misguided politicians from both parties (Collins from GOP, and Schumer from Democrats, for example) engaged in misguided dissing of the funding proposals as pork, so perhaps Brooks has a more general motive. Who knows? But, I think the piece is riddled with uninformed and misleading statements and may undermine public support for sensible efforts at infectious disease control.

Here is Brooks’ position in a nutshell:

You could say that the world should beef up the World Health Organization to give it the power to analyze the spread of the disease, decide when and where quarantines are necessary and organize a single global response.
…
The response to swine flu suggests that a decentralized approach is best. This crisis is only days old, yet we’ve already seen a bottom-up, highly aggressive response.

Brooks presents several arguments, all of which I think are all wrong.

In the first place, the decentralized approach is much faster…
…
The Times published a photo on Monday of the New York City health commissioner… leading a crisis response meeting. The photo is the very image of a focused, local response. People are wearing polo shirts and casual wear — intensely concentrating on the concrete incidents in their own backyard.

Brooks seems to have a fantasy notion of how newly emerging infectious diseases are combated, straight out of a TV show. Suppose that one day, big purple and yellow blotches broke out peoples’ on foreheads and then they blew up! New Disease! “Get on that case, Stat!!” the heroic resident yells, running breathlessly down the hallway, in gym shorts and flip-flops.

This outbreak dates from late March, when Mexican health authorities noticed that the there was an increase in cases at when there should have been a reduction if usual seasonal patterns had held. Does that seem ‘fast’? Or is does that timing have the slows from the dead hand of bureacracy?

Well considering that you cannot see a flu virus, and the symptoms of many influenza like diseases are similar, and only a few labs have the ability to detect new strains definitively, it was about as fast as can be expected. But, without a well organized reporting system, a network of calibrated labs with means to identify and classify viruses, we may still not even know there is a new bug out there. A report of the early cases is here:
http://www.who.int/csr/don/2009_04_24/en/index.html

Brooks continues:

If the response were coordinated by a global agency, those local officials would not be so empowered. Power would be wielded by officials from nations that are far away and emotionally aloof from ground zero. The institution would have to poll its members, negotiate internal differences and proceed, as all multinationals do, at the pace of the most recalcitrant stragglers.

This statement is just silly. The whole point of funding research and planning for pandemics is to iron out all these problems before hand. But, of course, if you don’t have the money to plan, coordinate, and for training exercises and drills before hand, then exactly what Brooks fears will occur.

…the decentralized approach is more credible. It is a fact of human nature that in times of crisis, people like to feel protected by one of their own. They will only trust people who share their historical experience, who understand their cultural assumptions about disease and the threat of outsiders and who have the legitimacy to make brutal choices.

This is also just silly. Does Brooks think that there is a flying shock corps of totalitarian WHO disease fighting Czars who parachute in to a country and do everything according to a uniform template? That might be the old IMF approach to macro-economics , but that is another matter. Of course, most primary care and country planning is done by local healthcare workers and scientists. His mention of brutal choices betrays more juvenile and uninformed TV-drama thinking. The whole point of preparation and early response is to prevent the necessity of making brutal choices whenever possible.

Finally, the decentralized approach has coped reasonably well with uncertainty. It is clear from the response, so far, that there is an informal network of scientists who have met over the years and come to certain shared understandings about things like quarantines and rates of infection. It is also clear that there is a ton they don’t understand.

In response to this, below are two URLs for the pandemic and epidemic preparedness pages, from the CDC and WHO, respectively. The reader can judge for themselves how ‘informal’ this effort arising from ‘shared understandings’ is.

Pandemicflu.gov
http://www.pandemicflu.gov

Epidemic and Pandemic Alert and Response (EPR)
http://www.who.int/csr/disease/en
(this page is a list of diseases covered, you can navigate up (to see overall program) or down (to explore issues specific to a disease)

Why all this bureaucracy? Well, I can think of a few reasons. You need clear guidelines for reporting possible new infections so the system isn’t swamped by false alarms. Infectious agents can be difficult to identify, and different labs may be appropriate for different circumstances (you want to maximize some combination of accuracy of identification and rapidity of response) –so you need guidelines on where to send which samples. Incorrect collection of preservation of samples may destroy them, so you need guidelines for those functions too. You need guidelines and standards for your own local labs for work done in country. I hope the reader gets the idea.

Brooks concludes:

Swine flu isn’t only a health emergency. It’s a test for how we’re going to organize the 21st century. Subsidiarity works best.

The OED defines subsidiarity as follows:

The quality of being subsidiary; spec. the principle that a central authority should have a subsidiary function, performing only those tasks which cannot be performed effectively at a more immediate or local level.

The reader can look through the CDC and WHO internet sites above, and judge for themselves whether this principle is followed now.

Some of Brooks arguments are misleading straw men. Others seem absurd on their face, but might appeal to people who do not think things through. Brooks favors the use of ‘emergent authorities.’ Do you think, when a potentially dangerous and possibly new infection is first detected, you want spontaneously "emergent authority" to figure out where to send it to be identified? I don’t think so, so you need a system, which takes money. Brooks also favors ‘experimentation’ Do you want experimentation when you are trying to ID the potential new bug? Collecting and preparing a sample? Designing a local disease monitoring system? I also do not think so.

Brooks comment on experimentation betrays a lack of understanding of non-experimental sciences, such as epidemiology, economics, and many kinds of biology and medicine. Reality, and your response to it is a continuing experiment. Careful planning, continuous evaluation of procedures, consultation with as many other authorities you can find, are essential. Even with all the planning in the world, and optimal mix of centalized and decentralized action and decision making, you will be thrown into the next real life experiment unexpectedly. To make arbitary and thoughtless decisions to let things just happen and hope for the best will guarantee a mess that must be cleaned up at great cost, and may provdie little information about the next emergency (that is, real life experiment) that will come your way.

A common finding in epidemiology and economics of infectious disease control, is that a global perspective is needed, which requires a coordinated reporting, testing and response system.

For example, it is often the case that there are two regimes in disease control –the ‘low disease prevalence’ and ‘high disease prevalence’ cases. Prevention and control is often most efficient if you can catch the disease in the low disease prevalence situation (when the disease is rare), then you have many options and it is practically feasible to economically efficient to keep the disease rare, and perhaps eradicate it. If you miss that chance, there is often no going back. Decentralized decision making, which feels its way along local solution without considering the global effects, cannot handle this situation. Two examples relevant to influenza like diseases are given below

To summarize, we showed that when designing control strategies for infectious disease outbreaks, it is not enough to consider a single outbreak. Instead, any comprehensive emergency preparedness planning also needs to consider how certain control approaches perform under a scenario where multiple outbreaks are possible.

Handel A, Longini IM, Rustom A What is the best control strategy for multiple infectious disease outbreaks? Proceedings of the Royal Society B (2007) 274, 833-837.

SARS (severe acute respiratory syndrome)…, emerged in late 2002 in China. …. Some scientists argued that this was not enough, however, that the opportunity to eradicate the disease should be seized before SARS had a chance to become established. As Burke (2003) put it, ‘epidemic-control efforts should not simply be maintained, but doubled, and redoubled again’. The epidemiological rationale for moving quickly was that there existed but a short window during which SARS could be readily distinguished from influenza. Wait too long, or act too passively, and eradication might cease to be feasible. This paper points to a further [economic rationale] rationale: While a short, sharp response may be optimal at the early stage of the disease, a sustained effort at eradication may not be optimal after the disease has become established.

Barrett S, Hoel M Optimal disease eradication Environment and Development Economics 12: 627–652

Brooks seems to fear inefficient, misguided centralized authoritarian control, and describes some of its dire effects of economic welfare and personal freedom –his column mentions possibility of ‘quarantines” and “brutal choices”. However, the government will always have the power to exert very drastic control on populations where dangerous infectious disease becomes common, and an unacceptable threat to life. His recipe for disease control, at least for diseases like influenza, make that dystopian vision more likely, not less.

A simple example is immunization. It is well known that there are significant externalities in immunization for influenza. I do not fear immunization, and recommend it for others. Some do not like it, and I want them to be able to do as they want. A well designed immunization program is more likely to protect the population, and permit those who do not believe to forgo it, without danger to their health. The situation where the government requires immunization to detriment of individual freedom will be less likely. Very often, centralized control and planning is required for an effective program that works, does more good than harm, and allows those who do not wish to be immnuzed to avoid serious harm from others’ actions. German measles, or Rubella, is an example.

Of couse, decentralized control of many diseases is optimal. Both economists, epidemiologists and doctors know that. Examples are sniffles and the common cold. Deadly vector borne diseases, like bubonic plague or cholera might be on the opposite end of the spectrum, where much centralized action is essential. Control and prevention of most infectious disease falls in a spectrum someplace in between. Where is influenza? There is evidence that the modern deadly flu pandemics often started as a relatively mild disease and then evolved in several successively more aggresive and deadly outbreaks in different populations in different countries, until they erupted in a deadly pandemic. Do you want the reasoning displayed in Brooks’ column to decide whether a decentralized ‘emergent authority’ of (informal?) ‘shared understanding’ deals with such a scenario? You be the judge.

Below is a review article written in plain english on the economics of infectious disease control for those who would like to read more.

Klein E, Laxminarayan R, Smith DL, Gilligan C. Economic incentives and mathematical models of disease. Environment and Development Economics 12: 707–732
http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=1343564
(may be firewall, so go to a library for free view -I can’t find a free copy)
The sections on externalities and global commons are especially relevant.