April 2006: Point/Counterpoint on Pay for Performance

Robert M. Centor, MD; Michael S. Barr, MD


April 19, 2006

In This Article

Point: Why I Fear Pay-for-Performance

On "Point," we welcome Robert M. Centor, MD, Professor and Director, Division of General Internal Medicine, University of Alabama School of Medicine; Associate Dean, Huntsville Regional Medical Campus; President-elect, Society of General Internal Medicine.

For every complex problem, there is a solution that is simple, neat, and wrong. -- H.L. Mencken

During the course of my medical career, I have experienced several solutions to the healthcare delivery problem. The big initiative of the 1980s and 1990s, managed care, was the solution to overspending and controlling healthcare expenditures. Managed care, while still present, has morphed into a system that no longer rewards primary care excellence. Few policy makers see managed care as the ideal method for providing our healthcare.

The most recent "fad" idea that we discussed involves quality and errors. We have commissions that decry the errors in medical care. The literature on quality continues to report on substandard quality in delivering evidence-based care.

The pay-for-performance movement (and does not the phrase evoke a sense of moral virtue?) assumes that physicians will provide better care if we provide financial incentives to do the right things. The concept has great validity on its face. Pay-for-performance has become a sound-bite phrase, which politicians eagerly adopt.

So who could oppose motherhood, apple pie, and quality? No one can oppose the drive for improved quality, but I do oppose current efforts to adopt pay-for-performance.

Excellent medical care requires excellence in at least 3 dimensions. First, one must make the correct diagnoses. If we expect correct treatments, we must assume diagnostic accuracy. Of course, difficulty with diagnosis ranges from trivial to very complex.

Second, one must deliver the appropriate care for an individual problem. If the patient has one problem, then an algorithm can direct quality care. I know that all patients who have congestive heart failure should have an angiotensin-converting enzyme inhibitor prescribed. However, we know less about how one should consider quality when patients have multiple diseases. In adult medicine, many patients have multiple diseases, each having complex care guidelines.

Third, we should develop a plan given the context of the patient's situation. We must understand the financial and social constraints of our patients. We must communicate with our patients, understanding who they are and what kind of care they desire.

I submit that current pay-for-performance plans only address part of one dimension of care. They will reward physicians for caring "correctly" for patients having a single known problem. But I also submit that this formulation may not reward the right physicians, or even encourage total excellence.

An example might help make this clear.

A 54-year-old man with type 2 diabetes mellitus goes to see his physician for a routine visit. The first physician checks his eyes and feet. He orders a hemoglobin A1c and a lipid profile. He adjusts the antihypertensive medications. Flu and pneumonia vaccines are administered. At the end of the 20-minute visit, as the doctor is leaving the room, the patient mentions that he has begun experiencing chest pain on his daily walks with his wife. The doctor, obviously in a rush to see his next patient, quickly reassures the patient that they will address that issue at the next visit. Two weeks later, the patient has a large anterior myocardial infarction.

The second physician starts by asking the patient if he has any new symptoms. The patient mentions the chest pain immediately, and the entire visit is spent on addressing the chest pain. She sends the patient for a stress test (which is positive) and then to cardiac catheterization. While doing this, she does not address the eyes, feet, glucose control, lipids, or blood pressure. The patient needs coronary artery bypass surgery.

If we just look at the chart for diabetes performance indicators, the first physician gets a perfect score. If we look at overall care, the second physician has done a better job.

Thus, one can imagine that focusing on one dimension of care could compete with adequate focus on other dimensions. Economists warn us to always consider the externalities of our decisions. Once we develop an incentive system, we will clearly prioritize how physicians spend their time with patients. These incentives might improve the care of some patients, yet decrease the time the physician spends on diagnosis. We might overemphasize filling out the "checklist" and ignore excellence in the doctor-patient relationship.

Even if we can avoid these externalities, we still have a problematic system. How do we balance multiple diseases? How do we prioritize treatment when constrained by economic realities?

Yet the appeal of pay-for-performance will persist because of its surface validity. I urge the medical community and payers to work to understand all the implications of pay-for-performance prior to advocating its adoption.

We must understand how a pay-for-performance system would influence patient care. What would happen to all the dimensions that play a role in overall care?

If we are not cautious, we will risk becoming like the drunk in the famous, oft-used joke:

A man sees his drunken friend circling a lamppost at night.
"What are you doing there, Bill?"
"I'm looking for my house key."
"But you lost it in the tavern, Bill."
"I know, but there's more light here."

If we develop an incentive system that only focuses on the light, then we may well miss other important aspects of doctoring. If so, patient care and outcomes may not improve.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.