Network News

X My Profile
View More Activity

Against Physician Pay-for-Performance

This guest post is from Sylvia Brandt, an assistant professor at the University of Massachusetts, Amherst.

During the Great Health Care Debate, a great deal of attention has been focused on the issue of physician incentives. Atul Gawande's article in the New Yorker on discrepancies in per-insured Medicare spending, which President Obama made required reading in the White House, highlighted the economic incentives that physicians have to order additional tests and procedures - especially when they can order those services from for-profit companies of which they are owners.

One proposed solution, reportedly favored by Peter Orszag and Obama, is to shift toward paying physicians for performance. The basic concept is simple: physicians' compensation would be linked to their patients' health outcomes, and therefore they would have the incentive to do what is most likely to produce a successful outcome at a reasonable cost. This idea seems obvious to many economists and policymakers. But when you look closely at the dynamics of illness, health care and household behavior, the picture becomes murkier.
Pay for performance faces some major drawbacks, especially where the treatment of chronic illness is concerned. Childhood asthma, an important chronic illness that has been the subject of my research for several years, is a good example.

From a medical perspective, asthma is straightforward to manage, and with correct management patients should rarely suffer any symptoms. However, correct management depends much less on the physician (although it is possible for a physician to prescribe an inappropriate management plan, such cases are relatively rare) than on the family - taking preventative medications appropriately, making risk-reducing investments such as air filters and mattress cases, avoiding risk-increasing activities and situations, etc. (The idea that people do not always do what they should do to take care of their health should be a surprise to no one; how many readers ate seven servings of fruits and vegetables yesterday?)

Whether a family takes these necessary steps depends not on the physician's instructions, but on that family's subjective perceptions of asthma and, perhaps most importantly, the resources available to it. It is one thing for a middle-class, suburban mother with a washer and dryer to wash her child's sheets on hot water weekly; it is another thing for a mother who has to walk to a laundromat in a unsafe neighborhood after a day of earning the minimum wage. In addition, for a large group of illnesses, it is increasingly clear that environmental factors beyond the control of families and physicians alike play a determining role. A child living next to a diesel bus terminal or rail station is likely to have worse respiratory health outcomes regardless of individual household behavior.

As a result, whether a child requires repeat hospitalization for an asthma exacerbation - something that should rarely happen, given proper medical care - is a function less of the behavior of the physician than of the behavior of the family and of environmental factors. This, in part, is what makes the problem of childhood asthma so difficult to solve; it is not simply a matter of inventing a new drug. In this context, a pay-for-performance plan could penalize physicians who are doing the best job they can. In addition, it would actually discourage physicians from serving those low-income populations that have the highest incidence of childhood asthma, precisely because they are less likely to be able to manage asthma appropriately (among other reasons, because caregivers have less time and money available). The end result could be even fewer medical services where they are needed most.

This problem is not confined to asthma, but is common to many chronic illnesses whose proper treatment is no mystery to modern medicine. Obesity, for example, has been estimated to produce $147 billion per year in direct health care costs, yet physicians have only limited influence on health outcomes. Of greater importance are neighborhood characteristics – such as parks or sidewalks - that make it easier or harder to get exercise. Diabetes is another example. For this class of common, widespread, expensive illnesses, pay for performance risks creating exactly the wrong incentives.

Many chronic illnesses, including asthma and diabetes, afflict poor people much more than the rest of society. Universal health insurance coverage will certainly help combat these illnesses, since it will at least ensure that people have access to the basic medical care that they need. But simply seeing a doctor is not enough; families need the money and time to make necessary health care investments – or to move to a non-polluted environment for their children.

The only comprehensive way of improving outcomes for people suffering from these illnesses is to combat poverty itself. Unfortunately, poverty seems to have slipped off of the national political agenda sometime in the 1980s. In the meantime, a simpler solution to the problem of physician incentives, at least for management of chronic illnesses, is to pay doctors a good salary, with additional financial assistance to repay student loans and therefore reduce the pressure on medical students to go into high-paying, procedure-driven specialties. This will eliminate the incentive to order unnecessary tests and procedures, which is one component of the health-care cost problem.

When it comes to the primary-care physicians who treat childhood asthma, my experience and research lead me to believe that they do not need any additional monetary incentive to do the best they can, and monetary incentives could have perverse consequences. For other types of medicine, of course, it is possible that pay for performance could have beneficial effects.

--Sylvia Brandt is an assistant professor in the Department of Resource Economics and the Center for Public Policy and Administration at the University of Massachusetts, Amherst. Her research focuses on the economics of health and household behavior.
Disclosure: Brandt is married to James Kwak.

By Terri Rupar  |  July 30, 2009; 1:42 PM ET
Categories:  Health care  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Attention Shifts Back to the Foreclosure Crisis
Next: Making Financial Regulation Work: The Supreme Court's Role


Pay for performance will just lead to MDs doing what insurance companies do now: refuse to treat the elderly, the obese, smokers, and cancer patients. Not a good outcome for ANY of those people!

Posted by: WashingtonDame | July 30, 2009 3:02 PM | Report abuse

Pay for performance puts doctors in the health insurance business, with all the perverse and destructive incentives of the insurance companies. We want doctors who take on the task of caring for the sickest, highest risk people who are most vulnerable to serious health problems---the poor, those in occupations with risk of illness and injury, those at risk of HIV, those who cannot easily educate themselves about healthful behaviors.

Pay for performance creates a financial incentive for doctors to seek out patients who have already done most of the work of staying healthy---the well educated, the higher income, the slim non-smokers who exercises and works behind a desk.

You will create a situation where, like with insurance companies, you pay physicians more to "cherry pick" the well, and avoiding those who are sick, at high risk of getting those pesky chronic illnesses, or unlikely to be able to manage those illnesses. Hello suburbs, goodbye inner city.

Pay for performance is essentially a gift to the insurance companies who will be delighted to download the financial risk of chronic illness to the most dedicated physicians who have chosen to care for the sickest and highest risk of our population.

Posted by: kenjaffe | July 30, 2009 6:08 PM | Report abuse

My father is a family physician. His medical center has tied physician pay to patient outcomes. The doctors made many of the same (justifiable) objections that you raise - the outcomes are dependent on the patient, rather than the doctor. But when the program was implemented, a funny thing happened - outcomes improved. So, no, the doctor doesn't have total control over the patient's outcome. But there are things they can do to help, and pay-for-performance would encourage this.

Posted by: pork1 | July 30, 2009 6:56 PM | Report abuse

Any compensation approach that relies on a single measure, even a broadly defined one like "performance," can probably be gamed fairly easily. The difficult challenge is to come up with a handful of measures, two or three at most, that taken together aren't easily manipulated. Risk-adjusted capitation payments could be one such approach, since they can take into account patients' ages, health status, and (conceivably) income status. Physicians would receive higher monthly payments for sicker and poorer patients, and could make better-than-average profits by improving these patients' health.

Posted by: wdarmes | July 30, 2009 10:34 PM | Report abuse

The posting is too simplistic. A reasonable system would take into account SES, age, and other indicators closely tied to outcome. I agree with 'pork1' and 'wdarmes' that a way can (and must) be found.

Posted by: charliecoop | July 30, 2009 10:42 PM | Report abuse

It's partly in the triage and partly in how you define performance. If triage is opportunistic, you'll obviously get a biased picture of effectiveness. Same if you measure performance by something that doesn't really get at treatment impact (like death probabilities unconditioned on accurate diagnoses). Wdarmes's risk-adjusted capitation payments sound like an answer. Do such things exist, and are they accurate and comprehensive enough?

Posted by: GingerBear | July 31, 2009 12:28 AM | Report abuse

Incentives. What is the incentive to be fat, to smoke, to never exercise? Does anyone enjoy being fat, huffing and puffing to do the most minor physical things? Does anyone enjoy the coughing that comes with smoking?

I live 2.5 miles from a discount grocery. I bike there to buy my food to save money. The place is well stocked with fruit and vegetables, lean meat, fish at prices 30 to 40% less than the mainline groceries (Safeway, Albertsons, Kroger). It is almost in the center of the poor part of town so the poor can walk there easily. When I shop I observe the poor pushing shopping carts filled with soda pop, snack foods, just about everything you could eat to hurt your health. I park my bike by itself because nobody bikes, instead the poor are loading up elderly vehicles with the junk food I mentioned.

What I want to know about is not incentives to help the poor have medical care regardless of bad habits, but of what incentives will get them to change the awful choices they are making, thereby saving themselves from medical expenses not to mention improving their quality of life.

If people are going into the water without knowing how to swim, should we spend money providing better life preservers or teaching them to swim?

Little by little the government assumes the role of protector and provider for more and more people who can remain ignorant and indolent because there is one kind of program or another for all to allow them to keep on keeping on.

All children must attend school and public schools are free in the United States. In his/her school years, doesn't every child receive instruction on healthy eating and healthy lifestyle? What does it cost to walk daily? What does it cost to eat less than you are if you are overweight? Can't a person look at their own waistline and see there is a problem?

In all the talk about health care coverage, I'm seeing little about personal responsibility and much about our obligation as a nation to take on this program regardless of individual cases. I don't get it. Could there be anything more likely to result in higher expenses for taxpayers that a health plan that puts incentives for doctors high in the priority list but pushes for little or no incentive for a change in poor behaviors by the millions of Americans the doctors see?

Posted by: Clif | August 2, 2009 9:20 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company