You want healthcare reform. I want healthcare reform. Grandma Jenkins wants healthcare reform.
What is “healthcare reform?” What kind of animal are we talking about? How would we recognize it if it came up and bit us? What are its markings, its behavior, its habits?
From observing the systems of other countries, from the results of local experiments and variations in the U.S. system, and from serious research over decades into outcomes and comparative effectiveness, we can actually outline what the marks of a better healthcare system would be.
But healthcare in the United States is a complex adaptive system. If we want to deeply and fully, we have to take one step back and revisit what we know about the nature of complex adaptive system, and how that knowledge might apply to reform of this system.
Healthcare is complex: It has many inputs and outputs, which operated independently upon one another in multiple overlapping feedback loops. Device manufacturers, for instance, adjust their costs and prices to reimbursement levels, and reimbursement levels are set to prevailing price structures. Preventive diabetes services, such as relatively inexpensive nutrition education, are under-compensated, and so are scarce; this leads to a need for more expensive services, such as emergency treatment of diabetic shock, and amputations.
All dynamic systems adapt continually. The various players (pharmaceutical companies, providers, health plans, consumers, employers, regulators, politicians) optimize their positions as much as they can with the resources they have access to (mostly money, but also other proxies for money, power, and positional security, such as votes, public sentiment, access to media, and systemic inertia). This is normal. This is how systems work.
This is also why our healthcare system, in almost universal judgment, is so dysfunctional. It has become optimized to the convenience and profit of the players with the greatest resources. All systems are in some sense self-righting: If the pikes eat up all the trout, then the pikes die off; without many pikes around, the trout proliferate until the pikes make a comeback, gorging on the trout. But in this case the healthcare system is dragging down the economy with its expense, and causing enormous personal economic misfortune, bankruptcy, misery, and death in the population. Waiting for it to right itself (or expecting that it will do so before causing ever-widening suffering and destruction) is a mug’s game.
The healthcare industry in the United States is, in game theory terms:
- Both competitive and cooperative
- Non-zero-sum – you don’t have to make others lose in order to “win”
- Infinite – with no end point, it is more like the stock market than football or chess
This “infinite game” has been a reasonably stable system, with each player performing their expected part (though often grumbling that they are not well served) because it has been, in game-theory terms, a near-perfect “Nash equilibrium,” a kind of strategic gridlock in which no player could benefit from any unilateral change in strategy, and in fact would usually be punished for it. A doctor who decided unilaterally to spend more time with each patient, a pharmaceutical company which unilaterally lowered its prices, even a hospital which managed to reduce its re-admit rate, or a hospital CEO who decided to forego a shiny new edifice and instead focus on re-engineering processes – all would be punished economically and professionally for doing what we, their ultimate customers, would like them to do.
However, the system is now showing symptoms of increasing instability, as various players perceive that they are doing so poorly at the game that a change in strategy might, in fact, benefit them. This includes doctors who opt out of the insurance payment system, or set up “concierge” practices, or open urgent-care centers; patients who go to foreign countries for care, buy pharmaceuticals over the Internet, or opt out of the medical system entirely because they can’t afford it; and hospitals like Geisinger who set up their own insurance system, hire doctors, bundle products, and give warranties. Players that show no little interest in major new strategies, such as pharmaceutical companies, health plans, and device manufacturers, are signaling that they feel that they are “winning” at the game as currently played – or at least that they feel that they are doing better than they would under any other strategy that they can see. Players attempting to quit the game or change the rules are signs that the game is breaking down.
The local optimization of players in a Nash equilibrium does not mean that the current strategic gridlock is actually the best for all concerned. There might well be some different configuration in which all parties are better off. But they can’t get there from here, without some interruption of the system from outside, some influx of new energy (like, for instance, new funding), some new players (like, say, a government-sponsored “safety net” insurance program), some shift in the resources of the existing players (like consumers or employers being given greater information and power to choose).
In practical, everyday terms, this point of view – seeing the healthcare as a complex adaptive system capable of analysis in terms of game theory – renders some useful observations and rules of thumb for evaluating any possible healthcare reform. They include:
- You get what you pay for (and the inverse, if you don’t pay for it, you don’t get it). Stick a scoop into the healthcare soup, and you’ll find dozens of examples, but here’s one: Give “pay for performance” bonuses for specific measures (number of diabetes patients getting eye exams, for instance) and that measure will improve. Other measures will not, and may in fact decline as resources are shifted to improving the specified measures. The assumption that PFP bonuses will cause a general increase in quality has proven generally unfounded.
- The Law of Unintended Consequences reigns supreme: To the closest approximation, all the most important consequences of any given scheme will be the unintended ones. Example: Charging customers co-pays. Intended consequence: Cut casual “over-utilization,” recreational surgeries, whine-on-demand hypochondriacal office visits. Actual consequence: Cut all minor utilization, including preventive checkups, pap smears, mammograms and so forth, thereby actually increasing major utilizations for the big things that the checkups didn’t catch; also cause some people to forego truly necessary treatment (chemotherapy, cardiac catheterization) and simply die rather than impoverish their families.
- Controlling specific costs and utilizations becomes a game of Whack-A-Mole. Example: Control length of stay and other in-patient cost structures, and suddenly you get lots of drive-through surgeries (“You want fries with that hip?”), until those come under control as well. Try to control pharmaceutical costs by refusing to reimburse for over-the-counter drugs, and suddenly there is a prescription version of ibuprofen, same stuff, just twice as strong, so that it can be reimbursed. This is the “adaptive” part of a “complex adaptive system.” The system perceives proscriptive regulation as damage and routes around it.
- Systemic decisions reflect the needs and desires of the individual decision-makers, not the system as a whole, or even the sectors within the system. If you want to understand hospitals’ strategic plans, for instance, you have to ask yourself how hospital CEOs make a living, what enhances their career prospects and what gives them more prestige and job security. The same is true of pharmaceutical company executives, doctors, health plan executives, consumers, legislators – anyone making a decision. Those needs and desires may line up with the needs of their sector, or with the needs of their customers or payers or constituents, or they may not. If they don’t, the needs of their sector or their community or their customers or constituents become just about perfectly irrelevant.
- Don’t expect anyone to “do the right thing.” They just won’t. It is close enough to the real case to say that they can’t, if they are punished for doing so. So don’t design any part of the system on the assumption that the various actors will “do the right thing.” Sure, in every profession there are people who swim upstream of the flood of incentives and do what is right by the people they ultimately serve, even to their own detriment. These people are heroes of healthcare. But heroes are rare, and their appearance is unpredictable. Any part of a system designed for heroes to step forward and sacrifice themselves will fail. In aggregate, expect the decision-makers in any sector to act in their own personal best interest.
This lesson has stood out vividly in the current financial crisis: Deregulators felt that bankers and other financiers would regulate their own behavior and do what would be prudent for their institution, their sector, and their customers. Instead, they fairly uniformly did what brought them the biggest salaries, stock options, and bonuses.
However obvious it is to an outsider what “the right thing” should be in another person’s situation, it is not at all obvious to that person. The surgeons doing the thousands of unhelpful spinal fusion surgeries, the doctors ordering the hundreds of thousands of unnecessary images, the health plans cutting off chemotherapy to people whom they have managed to re-define as “ineligible” – we can come up with lots of psychological and sociological characterizations of their motives. But the simplest explanatory principal is Upton Sinclair’s dictum: “It is difficult to get a man to understand something when his salary depends upon his not understanding it.”
There are probably many other rules of thumb that we could list here, but we could start with these. With a systems point of view in mind, we can turn to possible healthcare reforms and ask: What would be the markers of a healthcare system that would truly work?
By the same author: Fear and Loathing over the Stimulus Bill