5 August 2008 | by Presh Talwalkar | 8 Comments » |
Every Tuesday is a Game Theory article at Mind Your Decisions
Why are so many people taking drugs for cholesterol? I pondered this question after reading about a new study on Vytorin:
And the results left unanswered whether a drug that has been proved highly effective at reducing “bad” cholesterol—and has racked up $5 billion in sales in 2007 as a result—offers patients a proven benefit against cardiovascular disease.
[source: “More Vytorin Bad News Hits Merck, Shering.” The Wall Street Journal, July 22, 2008.]
The news is remarkable when you consider there are alternatives to drug therapy, such as diet and exercise. Dr. Dean Ornish showed heart disease can be reversed through lifestyle changes—and without use of lipid-lowering drugs—in a landmark study published in the Lancet in 1990 (reference). Other doctors have confirmed Ornish’s findings and even suggested we can use diet to become heart-attack proof. Here we stand 18 years later, and cholesterol-lowering drugs, not diet, are on the main stage.
What might explain the phenomenon? Are patients unwilling to change? Do drug companies have too much influence?
Perhaps, but then again, laziness and corruption are insufficient causes for two reasons. First, people do make diet changes. Just look at recent diet fads to see that people can and do make tremendous changes, if a big benefit is perceived. And second, doctors do care about patients and want to be ethical. Allegations about industry research have led doctors to take a stance on industry funding; some don’t want any.
I want to explore another explanation for our reliance on drugs. I hope this perspective will bring greater awareness and improve the situation for the leading cause of death.
I see the problem as one of incentives and hidden information, dependent on the strategic interaction of the doctor-patient relationship. I think the doctor-patient relationship in treating heart disease is much like relationship in another problem: the buyer-seller relationship in trying to buy a used car.
The market for lemons
Imagine you’re buying a used car using online car listings. You investigate the market and find listings ranging from $17,000 to $23,000. You’re in email contact with the owners and you’re trying to get the best deal. You’re participating in a marketplace with smart buyers just like you. Which car are you likely to buy?
You think strategically. At first, you consider buying the cheapest listing to get a good deal. But soon you worry that a lower price might indicate lower quality. After all, sellers know more about their cars than you do. They must know the car is worth even less. That thought scares you, so you consider looking at the more expensive listings in the hope of getting better quality. But you soon realize that some sellers prey on this attitude. They intentionally list a bad car at a high price to scam you.
You cannot estimate the quality of any particular car, so you have to base your buying decision on the average quality of cars in the market. If cars are equally likely to be good or bad, then you might decide it’s not worth paying more than the average price of cars that are listed. Initially cars are listed from $17,000 to $23,000, so you decide you won’t pay more than the average of $20,000.
But this price limit has a negative consequence. A seller of a high-quality car would not want to sell a car for such a low value, and hence the seller would be driven out of the market. In fact, all sellers who have cars valued over $20,000 would leave the market, as the following diagram illustrates.
Once these sellers leave, the market is limited to car listings from $17,000 to $20,000. But at this stage, you face the same problem of quality. You still can’t estimate quality, so you again have a problem with how much to spend. You again decide you won’t pay more than the average listing of $18,500 for a car. This move again drives sellers of high-quality cars out of the market.
The process continues over and over until the only sellers remaining have listings of $17,000. But there’s a problem: the only cars that would sell for the minimum price would be the worst-quality and worst-maintained cars, most likely lemons.
Most buyers don’t want these cars, and so no one buys or sells anything. The market completely halts.
(This example is based on a paper from George Akerloff in 1966 about the market for lemons. Akerloff’s idea also has applications to choosing a job and finding insurance.)
Doctors and patients
How does this relate to medicine?
Well, think about the following situation. After seeing a patient, a doctor chooses between recommending some combination of diet and drugs. For the sake of argument, assume that diet would work completely but requires patient motivation. Drugs provide a clear but lesser benefit and might have more risks. The major plus is that drugs require less patient motivation. The doctor’s goal is to give the most benefit to the most people.
Ideally the doctor could tailor a health program for each individual person. But the economics of managed health care and the mental limitations mean the doctor has limited time to make a decision. Typically, one decision will be preferred as a rule of thumb.
Which recommendation will succeed as the base?
Under these assumptions, the doctor might recommend diet to everyone that would follow it. If everyone complied, that would give the biggest improvements. The problem is if few complied, then patients would be at high risk and without drugs that could lessen the danger.
The issue is one of hidden information. It is not easy for a doctor to judge a patient’s motivation, just as it is difficult for a buyer to judge a used car’s quality.
In America, it might be appropriate to expect low patient motivation. Despite years of nutrition education, America is getting more and more overweight. To see how bad it is, you have to consider that the top 5 skinniest cities are just barely within the normal weight range. In the rest of the cities, the average person is overweight.
Just as car buyer is forced to use average quality, the doctor might consider that patients have average motivation. And it is this small factor that causes diet to lose out in the marketplace.
Step by step, recommendations are given based on patients with lower and lower motivations. The resulting guideline, based on the lowest motivated patients, or health “lemons,” is to give drugs:
Might such a market unraveling be happening with the treatment of heart disease?
Some solutions
Not all is lost. After all, used cars are still bought and sold—even high quality ones. The market for lemons is a theoretical concept, not necessarily a practical observation.
The used car market has developed protections to stop the market from unraveling. The main idea is to establish trust and improve signaling for high-quality products. Tools such as car histories, warranties, and a national reputation (like Carmax) all improve a buyer’s ability to judge quality.
Can the same ideas be applied for treating health problems? I hope they are coming and we can change our path of drugs. It is now joked in medical communities that the government needs to save people from themselves by fortifying water with medicine for cholesterol, blood pressure, and blood clotting. While tampering with our water is a joke, creating such a cocktail drug to help our health is not—it is a serious consideration for the worsening obesity epidemic.
I can see some solutions forming already. First, there are some doctors that have devoted their practice to lifestyle medicine. They create a separate market for highly motivated people. Second, patients are getting wiser about medication. There are many websites where you can find information on drugs and learn about alternative remedies. Talking to your doctor about these treatments can signal above average motivation. And finally, I heard there is a push to improve the doctor-patient trust through better communication, possibly web based. Just as the web has changed car buying, it might also revolutionize the doctor-patient relationship.
8 Comments » |
