December 5, 2005
Paying Doctors More Changes Behavior…
But does it lead to better care? And who defines “better”?
P4P. Pay for Performance. What is it? What does it mean? And why are we doing this anyway? Could the money be spent more productively?
The idea of paying all physicians the same amount for a service no matter the quality of service and judgment, the total cost to the system, or the quality of controllable outcomes seems ridiculous. Yet it is the way the government and commercial insurers have done it for generations, creating as a side effect the fabrication that all physicians are fungible.
This approach is predicated on the fallacious assumption that an MD degree somehow confers equal quality to the service delivered. Most assuredly not true.
So-called Pay for Performance in theory could make some sense. To do this most effectively would require a true market interaction between the customer and the provider of service. The market--and not the government, insurers, or panels of “experts”--would define quality, price, and value. Then the millions of transactions per month over time would determine true pay for performance.
But alas, we have a government; we have commercial insurers; and we have enough “experts” to gag a maggot. And they all, just like physicians, think they know what is right. Instead of true outcomes, they are devising a system of process measures as surrogates for the quality of true personal outcomes.
This implies that someone external to the delivery and use of the service can sit in a room and study regression modeled statistics and say what constitutes “quality” for the customer. Not gonna work.
Enter “Evidence Based Medicine.” This movement has as its well-intentioned goal to define what constitutes “correct” medical practice for a given disease. Mostly this is being applied to the care of chronic disease. On the surface it makes sense. Dig deeper and you rapidly run into a huge Uh-oh.
Enter “Induction.” The problem is, we want to simplify things so they can “work” by taking deductively derived principles and applying them to individuals who may not fit the definitions. All medical studies are done on well-defined and highly selected populations in an effort to control as many variables as possible. Then the population is studied; usually regression analysis of some sort is applied to make statements about the validity and applicability of the findings to another population; and we create a fantasy that somehow population “statistical” significance is clinically significant to the next patient. Which leads to the question of how does cardiac risk in Framingham, MA bear upon my patients in Laredo, TX ?
A quick example: one time in my life I was a member of the disease management police. We were applying “evidence based medicine” principles to the care of a population of patients with diabetes and cardiovascular disease. When a patient was not on the “right treatment” as defined by the guidelines, it was my job to call the patient's physician and find out why and to vaguely threaten him or her to do what the protocol said.
What I rapidly learned was the “variations” we encountered were almost always known to the physician. There was almost always a good reason why the physician and/or the patient had decided not to use the “evidence-based medicine” medicine or treatment. A memorable example was the 85-year-old woman who decided against taking a statin medication for minimally elevated cholesterol because she failed to see the benefit relative to the risks and cost—a decision derived in collaboration with her physician.
Should she or not? There aren't good data in otherwise medically stable 85 year olds with minimally high cholesterol--who are normally not part of studies--to say what is “medically right” much less what addresses her own preferences. And I rapidly concluded that the only reason I cared was I was trying to make our outcome statistics look better. Phooey!!! Found a better job elsewhere.
A couple weeks ago the Robert Wood Johnson Foundation published the results of a three year study looking at seven P4P projects. Some have declared partial victory by defining “improved quality” as physicians who collect and record more data and physicians who can tick off the boxes of “evidence based medicine” protocols.
So, if you pay physicians to collect and record data and you pay them to tick off boxes, they will in fact do that. This is actually what the study says. It is a giant leap to conclude that better care has been delivered. And if this is management of chronic disease, why do we need physicians for this?
Because we don't have any good way to measure true personal quality and because we rely on governments, insurers, and “experts,” we have to resort to surrogates of personal value and quality. Perhaps the most absurd is paying physicians to collect and record data. We now employ the most highly skilled and highly paid person in the system to collect and record data, the least value add of any intellectual activity.
The outcomes are largely process ones. For example, we can show that aspirins get given, physician visits occur, and wellness tests get done. We have no data that this activity results in better personal outcomes or even population ones at this point. We have made a giant assumption that uncritically doing certain things will ensure better outcomes. So far, what we have shown is we can induce providers to change their behavior with money.
I think we already knew this. Big pharma has known for years you can change physician behavior with a pen. Not exactly what folks had in mind with the “power of the pen.”
I suppose this is a start on the quality issue, but completely unnecessary, complex, and, I suspect, doomed to collapse under its own weight eventually. What would make more sense is to continue to move in a direction where the finance and delivery of healthcare are separated.
And where the financial transaction and service transaction occur between the same two people.
Then and only then will we see price decline while quality and access improve.
Robert B. Teague is a pulmonologist and business consultant who is based in Houston, Texas. E-mail him.
Read other blogs in this series.
