stimulant

I’ve been thinking about how to design self-suffcient, scalable communities. A big theme in my thinking is that scale is a fundamental, but often overlooked factor in the behavior of complex adaptive systems (CASs). We don’t have a good handle on any of the basic, conceptual primitives we want to use when discussing CASs: complexity, order, disorder, structure, patterns, emergence — none of these linchpins are well-articulated, much less well-understood. However, we have made significant inroads understanding what scale means, and the roles that it plays in the natural world.

Despite the fact that the science of networks and complexity is pretty stylish, the really valuable ideas have been slow to percolate outward from the natural sciences. This is understandable, given that scientists are confused about all those other big ideas — when these get bandied about, our ignorance is rarely part of the qualification.

This is a real shame, because I’m increasingly convinced that simple ideas of scale can be very revealingly employed in thinking about social and societal reform. More specifically, many of the problems that we see in our social organizations are a product of mismatched scales, i.e. trying to solve problems on the wrong scale. To generalize, our society too frequently implements top-down solutions that attempt to address problems at a very high scale, using centralized, hierarchical infrastructures which are intrinsically brittle.

But, let’s start from the beginning with the case in point I’ve been thinking about recently: health care.

Ivan Illich’s Tools for Conviviality introduced me to the idea that modern medicine has stolen from us a sense of ownership of our bodies. We’ve lost our natural authority when it comes to our health. This has been replaced by a slavish attachment to the opinions of doctors. In Illich’s words:

The crisis of medicine lies on a much deeper level than its symptoms reveal and is consistent with the present crisis of all industrial institutions. It results from the development of a professional complex supported and exhorted by society to provide increasingly “better” health, and from the willingness of clients to serve as guinea pigs in this vain experiment. People have lost the right to declare themselves sick; society now accepts their claims to sickness only after certification by medical bureaucrats.

The same way that students are unable to gauge if they’ve learned something without a teacher, we’re unable to determine whether we’re healthy or sick (or how we’re sick or how to get better) without a doctor. This dovetails surprisingly well with how poorly studies revealing the aggregate irrelevance of today’s hypermedical health care have been received:

So I want to say loudly and clearly what has yet to be said loudly and clearly enough: In the aggregate, variations in medical spending usually show no statistically significant medical effect on health. (At least they do not in studies with enough good controls.) It has long been nearly a consensus among those who have reviewed the relevant studies that differences in aggregate medical spending show little relation to differences in health, compared to other factors like exercise or diet. I not only want to make this point clearly; I want to dare other health policy experts to either publicly agree or disagree with this claim and its apparent policy implications.

Robin Hanson first alerted me to this via the wonderful blog Overcoming Bias:

Readers of the medical literature, as well as readers of medical media coverage and students in health and medicine, have all been given the strong impression that in the aggregate, more medicine produces more health. This was the impression thirty years ago as much as today. Yet our single clearest data point, the RAND experiment, confirms the typical result of aggregate correlation studies: we see no such relation. Thus the medical research literature must suffer from severe biases, such as fraud, funding bias, treatment selection bias, publication selection bias, leaky placebo effects, misapplied statistics, and so on. How else can we square the usual positive benefit found in medical publications with a net zero benefit? Furthermore, what else but education and media biases can explain why this experiment, very expensive, well published, and the most important medical study ever, remains mostly unknown to medical students, professionals and the public?

<

p> This prompted me to start thinking about health care more carefully. The first thing I do when confronting a systemic or infrastructural problem is to consider the role of scale. Scale is a factor which complements Illich’s statement about stolen engagement, when it comes to both learning and medicine: when we try to address these problems as societal factors, we naturally use a top-down approach. Top-down approaches are well-known for their tendency to disempower, disenfranchise, and dehumanize. There aresituations wherein they are desirable. Top-down solutions are very good for top-down problems. But generally — particularly as the internet makes emergent solutions increasingly flexible and feasible — bottom-up solutions are more robust, effective, and reliable. What does this mean for health care? First of all, given that in aggregate, medical spending is overwhelmingly ineffective, the fact that the largest consumers of health care are those who benefit the least from it (i.e. the elderly), it’s clear that more emphasis on lifestyle as opposed to health care is desired. As Longman points out in The Health of Nations:

To get an idea of how wildly ineffective our health-care system is, consider this: The United States spends roughly $4,500 per person on health care each year. Costa Rica spends just $273. That small Central American country also has half as many doctors per capita as the United States. Yet the life expectancy of the average Costa Rican is virtually the same as the average American’s: 76.1 years. How can that be? According to public health researchers, the biggest reasons are behavior and environment. Costa Ricans consume about half as many cigarettes per person as we do. Not surprisingly,they are four times less likely to die of lung cancer. The car ownership rate in Costa Rica is a fraction of what it is in the United States. That not only means that fewer Costa Ricans die in auto accidents, but that they do a lot more walking, and hence they get more exercise. Thanks to a much lower McDonald’s-to-citizen ratio, the average Costa Rican thrives on a traditional diet of rice,beans, fruits, vegetables, and a moderate amount of fried food — and therefore enjoys one of the world’s lowest rates of heart disease and other stress-related illnesses.

More importantly, the way we dispense all this unnecessary care is itself broken. Although it’s unclear whether alienation from our body is the chicken or the egg (vis a vis the exaggerated importance of medical care), it is clear that the problems of health care are essentially local in infrastructure. There is a significant mismatch between the scales of the system currently in place and the set of problems the health care system alleges it addresses. The difficulty in creating a distributed system is that the current infrastructure has created a very insular, expert class whose centralization unsurprisingly recapitulates the same, structural problems.

The scarcity of expertise puts constraints on the scale of the system, and it seems that these constraints result in the overly top-down approach, yielding the mismatch in scale I blame for a lot of the health care system’s problems. But what if we question how atomic a doctor’s expertise actually is: the granularity of medical knowledge is much finer than “seven years of medical training.” In reality, medical knowledge can come in bits and pieces as small as knowing how to set a broken bone. The problem of health care is really the problem of distributing this resource (medical knowledge) and effectively pairing it with the materials necessary to deploy this knowledge (e.g. medical equipment).

Consider what a medical system that takes advantage of this granularity to create an emergent, distributed health care system might look like. There could be public x-ray machines which people could become licensed to use. These people would be indexed and reachable, and you would create a “micromarket” for their skills. Some people would only occasionally render their services, others would make a living providing high quality medical service. There would be a distribution of medical knowledge in the general population ranging from novice to expert, and people could capitalize on this knowledge as necessary. You could find and call upon neighbors specializing in what you need. It would become not only prudent, but natural to be engaged with your body and your health. Your medical identity would travel with you, not with your doctors. Your medical history and background would be generated by your use of this system, and this would provide the foundation upon which doctors could build, with their particular expertise, when necessary. But it is certainly clear that most visits to the hospital are for common, curable, simple medical problems. Why not re-empower people to solve their own problems, and those of their neighbors? Patients could rate the services provided by various groups and people, giving rise to a natural market [pressure] to sustain the sector. Lab work requiring equipment not feasibly made public can be outsourced — the same way you can have t-shirts printed today, you could have your DNA sequenced and your blood tested. This would create enormous demand for high-throughput, high-quality labs within a community. Response time in disaster situations would be instantaneous. The system would promote awareness and militate against the exploitation of people’s current insecurity and ignorance when it comes to health. The insurance landscape would be completely turned around: as you put more power in people’s hands, you put more responsibility, as well. The rarity of doctors currently perverts market pressure, but by distributing expertise, you can distribute not only the time load, but the vacuum in expertise across millions of people. The leading cause of malpractice isn’t incompetence: it’s the fact that doctors and nurses are rushed and overworked because the health care system isn’t keeping up with our demands. Decentralizing it changes not only the supply, but fundamentally alters the way that demand for health care is created. The benefits just keep piling up — you can imagine plenty of cute and clever subsystems to make this run smoothly, effectively, and far better than what we have currently.

I ran across a story on kottke.org about Jay Parkinson today:

Dr. Jay Parkinson M.D. emailed in to tell me about his new medical practice in Williamsburg. He’s got no office (housecalls only), takes appointment requests via SMS, email, or IM, handles some follow-ups over video chat, and specializes in the 18-40 age group without traditional health insurance. The goal, states Parkinson, is to “mix the service of an old-time, small town doctor with the latest technology to keep you and your bank account healthy”.

The infrastructure exists to support a vision of distributed health care! It’s happening already! The market possibilities are endless. Rather than tying up services with the medical infrastructure, you tie services to their only real requirement: expertise. This expertise then shares communal medical resources (e.g. public x-ray machines). Streamlining the routes to distribute the basic resources medical care provides (expertise, medicine, and equipment) is simple when you envision what it means to decentralize the distribution.

I’m eager for someone to critique these ideas — scale and decentralization seem to frequently fall just short of a panacea. Someone pinch me, tell me what’s wrong.