Sunday, March 13, 2005
Cutler's approach is radically different. He says that most health-care spending is actually good. Spending has been rising, he says, because it delivers positive, and measurable, economic value, and because it can do more things that Americans want. Therefore, Cutler says, we should focus on improving the quality of care rather than on reducing our consumption of it. Rather than pay less, he wants to pay more wisely -- to encourage health-care providers to do more of what they should and less of what is wasteful.
What is it that we do that is wasteful? Well, we probably prescribe antibiotics for viral illnesses more than we need to. Get back x-rays for pulled muscles, ankle x-rays for sprained ankles, and CT scans and MRI's for headaches more than we need to. We probably order more blood work than necessary, especially when someone gets admitted to the hospital or the emergency room. We admit more chest pain than is necessary, too. And we probably prescribe more Viagra and Prozac than is really necessary - at the behest of the patient. Sometimes this is in the service of defensive medicine, but often it's also done to provide much needed reassurance to an anxious patient who "just needs to be certain there's nothing wrong." And, of course, in the case of lifestyle drugs, there's a marketing factor at work that pumps up patient demand. It's hard to say no to a drug for which there is no objective measure of need. That's why, for certain things, the market approach, or "ownership approach" to medical care makes sense. When a patient has to bear at least some of the cost of his care, he's more likely to believe the doctor when he says a test or a drug isn't really necessary.
It's for catastrophic, emergency, and chronic health problems where the market solution runs into problems. And it's here that the pay-for-performance approach hopes to make a difference. For chronic diseases, there are certain easily measurable parameters that we know to make a difference in outcome. Diabetics who keep their blood sugar below 120, for example, do better than those who run in the 200's. Hypertensives who keep their blood pressure below 130/80 do better than those who run around 150/90. People with heart disease who take aspirin and keep their cholesterol below a certain level do better than those who don't. And those are the kinds of things the new Medicare reimbursement system considers:
Participating practices will evaluate 32 ambulatory care measures that focus on common chronic illnesses and preventive services. Among them: HbA1c management and control, lipid measurement, and eye and foot exams for diabetes management; left ventricular function assessment and ejection fraction testing, blood pressure screening, beta blocker and ACE inhibitor therapy for congestive heart failure; antiplatelet therapy, blood pressure screening, lipid profiles and low-density lipoprotein cholesterol levels for coronary artery disease; and blood pressure screening and control and breast cancer and colorectal cancer screening for preventive care.
In theory, this is a good idea. However, the reality of it is often very different. For example, I have several insurance companies who provide me with regular feedback on my "quality of care." Some of them provide me with data based on all of their enrollees who are my patients. This is useful. But others only look at one enrollee. Needless to say, if that enrollee happens to be a non-compliant patient, it makes me look bad. I may have tried every diabetic medication under the sun, sent my patient to diabetic education classes, and to an endocrinologist, but if his HbA1c level is subpar, I'm still the bad guy. They don't consider the effort, only the outcome. And the insurance company improves the chances of a bad result for me by limiting the sample size. (It's a good result for them. No bonus payments for quality!)
Which brings us to the other problem with pay-for-performance. Proper medical care of chronic conditions requires two willing people - the doctor and the patient. If the patient won't cooperate, there's nothing the doctor can really do. Cutler realizes this, but can't seem to admit how important the patient's free choice is in the equation:
''My grandmother used to say, 'Why should I take the blue pill -- I'm feeling better,''' Cutler recounts. ''The blue pill was why she was feeling better.'' As Cutler told me in the first of our many conversations, ''Ultimately, every discussion of health care turns personal.'' .
....This is especially true for chronic patients, whose well-being depends on following a long-term regimen of care. Diabetics, for instance, should receive yearly eye exams, regularly monitor blood sugar and cholesterol and take other steps to avoid problematic (and expensive) complications. ''Doctors say, 'You really should get your eyes examined,''' Cutler notes. ''There is no follow-up. Every doctor you talk to says: 'I know we don't do a good job on that. We don't get paid for it.' My way, we would pay them.''
The successful examples in the article are mostly from hospitals, where patients are captives with no choice but to take their medicine; where no smoking is allowed, and where even the number of calories ingested can be controlled. It's different in the real world, where a patient has freedom of choice. The real world is where a patient is more likely to take the "little purple pill" because it relieves his chronic indigestion, rather than the "little white pill" that silently controls his blood pressure; where coffee, and chocolate, and cigarettes beckon; and where people's bodies do not always respond perfectly to our interventions.
The pay for performance model also completely ignores the customized approach to treating a patient in favor of a cook-book-numbers-only approach. You can pump a person full of drugs to get his blood pressure down to the desirable goals, but he may be too sick to move as a result. And you can pump a person full of diabetes medications to get his HbA1c to goal, but he may have episodes of hypoglycemia that leave them comatose. But that's not important in the pay for performance model. It's the blood pressure numbers and the blood sugar values that will determine payment, not the patient's overall well-being.
In the long-run, what will happen to those patients who can't achieve numerical perfection? Or who can't quit smoking? Or who just can't bring themselves to have a colonoscopy? Will doctors refuse to take them on as patients, knowing they won't get paid for taking care of them? It would be a very mercenary thing to do, but it's likely to happen. A doctor who can't pay his bills is a doctor who can't practice medicine.
I know at least one of my diabetic patients would fair even worse than she does now under the system of the future. She refused to check her blood sugar, claiming it would just make her worry all the time if she knew what it was. Half the time, she wouldn't take her medication. She wouldn't see a nutritionist. According to her sister, she ate chocolates by the bagful. She wouldn't get her cholesterol checked because she had to fast beforehand. I could only get her blood work by having my staff draw it while she was in the office. She wouldn't see an eye doctor.
After six years of coaxing, I finally persuaded her to go to diabetes education classes. She still won't get a fasting cholesterol panel, but she sees an eye doctor now and checks her blood sugar once a day. She still eats chocolate by the bagful, but at least it's sugarless. Her blood sugar control isn't perfect, but it's a lot better than it was. And who knows, in another six years we might achieve perfection. Would I, or anyone else, have given her that many chances in a pay for performance system? Of course not. And that, ultimately, is the most disturbing thing about the "pay for performance" formula - its failure to acknowledge, or tolerate, our very human imperfections.
posted by Sydney on 3/13/2005 04:29:00 PM 1 comments
Refer to Diabetes for
By 3:55 AM, at