THOMAS H. LEE

A CAREER IN MEDICINE HAS ALWAYS COMBINED intellectual challenges with altruistic goals. These attractions remain intact, and the good news is that physicians today can do more for their patients than at any time in the past because of tremendous scientific progress. The bad news is that this same progress leads to chaos, inefficiency, and disappointing quality of care when imposed on a fragmented, disorganized healthcare delivery system.

The role of physician leaders is to bring order to that chaos. In my organization, we tell physicians that we must undergo two revolutions:

An industrial revolution, in which we develop and implement systems that improve care, such as electronic medical records with decision support software that helps clinicians make care safer, more reliable, and more efficient.

A cultural revolution, in which clinicians recognize that they are members of teams, and that these teams have responsibility for improving the care of populations of patients over time – not just reacting to the problems of the patients while they are sitting or lying in front of them.

That all sounds well and good, but revolutions tend to be miserable experiences for the people living through them. Accordingly, physician leaders need expertise in articulating the need for change, and moving that change along. A few of the basic tasks that dominate the lives of many physician leaders are:

Making choices among a range of beneficial options that are collectively not affordable

Creating financial and other types of incentives that encourage other physicians to improve quality and efficiency

Develop collaboration among parties whose interests are not in alignment

The tools needed to address these issues are rarely taught in medical school. But a good introduction to them can be found in economics courses and texts.

Making Choices

When they are working, physicians tend to be focused on one issue – helping the patient in front of them. Their patients like that attitude. Both physicians and patients would like to believe that resources are unlimited — even though rarely a day passes without a story about how rising healthcare costs are threatening the U.S. economy and blocking 47 million Americans from access to health insurance.

Until recently, many physicians believed that the challenge of making the entire healthcare system work financially was not their problem. Their obligation was to do anything and everything to help their patients. A common refrain was “If a treatment costs a million dollars, but it extends my patient’s life a little, or makes them feel a bit better, my duty is to give that treatment.”

There are still physicians in practice who adhere to this belief, but medicine’s leadership now supports a more complex approach to medical ethics – one that says physicians also have roles as trustees for the healthcare system. In short, physicians have duties to their patients, but they also have duties to the rest of society that require efficient use of resources to maximize their overall benefit.

Once one gets past the ethical and policy discussions, the real work begins – making tradeoffs in a world with constrained resources. Economics as a discipline provides tools for thinking about choices that bring different amounts of value to different people. Terms like opportunity cost and moral hazard were known by few in health care a generation ago. Today, they are part of everyday conversation for physicians in leadership roles.

Incentives

Getting physicians to change how they do their jobs is brutal work and not because physicians are bad people. In general, I have found the opposite to be true. They are mostly wonderful people who work hard on behalf of their patients. But they are also unbearably busy people who struggle just to get through the day without making mistakes or letting someone down.

As a result, they are not too excited about taking the time to learn new skills – like how to use an electronic medical record. They do not like to look or feel foolish, as might happen when it becomes clear they do not know how to type or use a mouse. And they hate even the slightest delay introduced into their day by computerized reminders to use Drug B instead of Drug A because of cost considerations.

To get physicians to do what they do not naturally want to do, their leaders have to use incentives. Peer pressure (“Everyone else is doing this”) helps, but generally only leads to partial success. To create major change reasonably quickly, financial incentives must usually be used.

There is more science to the use of incentives than offering more money for certain tasks or outcomes. The kind of questions that physician leaders must consider include: What potential unintended perverse consequences might occur? How much money should be offered as an incentive? Should the incentive be “packaged” as a bonus or potential penalty?

Useful insights on how to use incentives come from behavioral economics advances such as Prospect Theory, which won the Nobel Prize for Economics in 2002. Described by Daniel Kahneman and Amos Tversky in 1979 (1), Prospect theory describes how people evaluate potential losses and gains. The classic asymmetric S-curve of prospect theory displays actual financial gains and losses on the X axis, and perceived value on the Y axis.

Two key implications of Prospect theory are:

Small financial incentives can stimulate change more efficiently than large financial incentives. If you give someone $1000, they will be happy. If you give them $2000, they will be happier, but they will not be twice as happy. Therefore, if you have $2000 available in incentive funds, you are likely to get greater action by offering two separate $1000 performance incentives than one $2000 target.

The threat of losses has more emotional impact than the potential for bonuses. As the curve in Figure 1 shows, a $1000 bonus will lead to higher “value”. But a $1000 loss will lead to an even greater change in value in the opposite direction. Therefore, the most effective incentive programs threaten people with losses rather than promise them potential rewards.

Prospect Theory suggests that the “sweet spot” where incentive dollars can produce the most response is to the left and close to zero — i.e., small amounts of money packaged as potential losses. For example, physicians can be told that their salary is $150,000, but if they hit targets A, B, and C, they will receive bonuses of $2,000 for each goal. Or they can be told that their pay will be $156,000, but if they fail to reach targets A, B, and C, they will lose $2,000 for each. Prospect Theory and the author’s experience indicate that the latter approach is more effective.

Such incentive programs are, of course, irritating to the people at whom they are directed for the very reasons that they produce more action. Therefore, physician leaders need to have a sense of how their colleagues are responding to incentive programs – if they wish to remain physician leaders.

Promoting Collaboration

Why can’t people work together collaboratively even when their long term interests are really aligned? Some insight into this everyday phenomenon in medicine is provided by Game Theory, an offshoot of economics that was developed in the Cold War era by mathematicians interested in how rational but distrusting parties interact. In strategic games devised by game theorists, they studied the strategies that decision-makers use to optimize their outcomes – regardless of the impact of their decisions on others.

The classic example of Game Theory is Prisoners’ Dilemma. Two criminals (Prisoners A and B) are arrested, and interrogated in separate rooms. Both know that if neither betrays they other, they will both receive light sentences. They also know that if Prisoner A betrays B, A will go free, while B will receive a long sentence (and vice versa). The consistent result: both prisoners betray the other and receive long sentences, because they cannot trust their former partner.

Game Theory is intellectually interesting and fun, but carries a depressing message: rational people find it hard to trust each other, and therefore have a natural inclination not to cooperate. But how then do people sometimes find a way to cooperate despite this tendency?

This question was the focus of The Evolution of Cooperation, a 1984 book by Robert Axelrod (2), and a classic 1981 article of the same title by Axelrod and the late evolutionary biologist William D. Hamilton (3). In this work, Axelrod describes how cooperation develops despite the tendencies that discourage it, and what strategies are most likely to lead to cooperation most quickly.

Axelrod begins this examination with a description of spontaneous instances of cooperation that occurred in the trenches during World War I – the most famous example of which was a Christmas Eve soccer game between British and German troops. The actual extent of cooperation went much further. For example, reflecting the “Live and Let Live” philosophy that evolved, troops on one side would shell the other side with mortars, but would do so on a rigid schedule, and aim for a specific point in the enemy’s trenches. This predictability allowed the other side to minimize casualties. The favor was returned when the enemy used their artillery in the same way. The leaders on both sides were satisfied that shelling was occurring, while the soldiers in the trenches found a way to protect themselves.

Axlerod explored the development of such cooperation by conducting repeated games of Prisoners Dilemma, testing various strategies. The approach that led most quickly and reliably to cooperation was the simplest strategy, Tit for Tat. In Tit for Tat, one should never be the first to “defect” or do anything to harm the other side. However, if the other side harms you, you should not ignore it – you should retaliate. If the other side takes aggressive action again, then you should retaliate again. But if the other side then offers an olive branch, you should respond in kind.

Axelrod lays out a set of recommendations for how to pursue collaboration that go beyond Tit for Tat. He encourages the use of frequent face-to-face meetings to help both sides understand that they are “stuck” together, much as the soldiers in the World War I trenches were. More difficult is his principle that neither side should be jealous – that is, people should not dwell on whether the other side is getting a better deal, but instead focus on whether they are better off than they would be otherwise.

In conclusion, leadership in healthcare is needed more than at any time in the past, and physician leaders need skills beyond knowledge of clinical medicine. They need to understand how to make difficult choices in a world of constrained resources, and how to engage their colleagues and the rest of society in reaching consensus on these tradeoffs. They need to know how to use incentives to encourage other physicians to take time from their busy lives to learn new skills and adopt systems that often slow them down. And they need to know how to negotiate and encourage warring constituencies to cooperate. These skills are not now routinely taught in medical school, but perhaps they should be. In the meantime, one can begin to acquire them in economics courses and texts.

Thomas H. Lee, M.D. is currently the Network President for Partners Healthcare System. He also serves as a Professor of Medicine at Harvard Medical School