« Taxation and fairness | Main | China Fact of the Day »
Live, or Die Free
Johnson & Johnson has proposed that Britain’s national health service pay for the cancer drug Velcade, but only for people who benefit from the medicine, which can cost $48,000 a patient. The company would refund any money spent on patients whose tumors do not shrink sufficiently after a trial treatment.
The groundbreaking proposal, along with less radical pricing experiments in this country and overseas, may signal the pharmaceutical industry’s willingness to edge toward a new pay-for-performance paradigm — in which a drug’s price would be based on how well it worked, and might be adjusted up or down as new evidence came in.
More here. Contingency fees for doctors and pharmaceutical companies are a very good idea (one I have long supported). For more see Hyman and Silver's excellent paper.
Posted by Alex Tabarrok on July 16, 2007 at 07:21 AM in Economics, Medicine | Permalink
Comments
Very intriguing idea.
Posted by: Chris Meisenzahl at Jul 16, 2007 7:38:51 AM
Very Interesting, but sadly Healthcare is a matter of life and death and not of chances. The paitient could die due to wrong medication and if the medication does not take affect on the paitient then it will be considered wrong medication. Would Johnson and Johnson be willing to
1. Pay back for the time
2. Loss of life (if it happens)
3. Pay back for other medical expenses
4. Pay Other Damages
5. What about the hell that the paitients family has to go through.
I agree that the Pay-for-performance paradigm is a great thing to move to in the new world, but for some things like, Healthcare, Security, Law or where there is a matter of Life and Death it may not be applied.
WE have seen law firms charge more and more premium if they are known to solve / help accused. Look at India, we have law firms who specialise only in saving people blamed for car accidents (hit and run) and other typical criminal activities. At a high premium these law firms will make sure you don't end up in Jail, but then, just because the other lawyer was not competent enough, but his client was innocent, we don't want him to be in Jail too.
GS
Posted by: Gagandeep S. Sapra at Jul 16, 2007 8:29:10 AM
This is quite possibly the best blog post title ever.
Posted by: Neel Krishnaswami at Jul 16, 2007 11:45:56 AM
Discussing all the issues raised by Hyman & Silver could fill a book, but let me introduce a couple of key points:
First, the examples of successful contingency fees (or results-based compensation arrangement, RBCA, as they call it) given in other fields include legal contingencies, real estate commissions, and sales bonuses. The key difference between these arrangements and health care is that the former are splitting a easily quantifiable positive return on the service provided. The results of health care are either not easily quantified, e.g., the value of potential years of good health, or are the prevention of a negative 'return', e.g., death or disability, or both. This implies that RBCAs will be best suited for 'enhancing' therapies with a quickly determined outcome, e.g., cosmetic surgery, pain relief, sports performance enhancement, etc.
The issue of prevention (or amelioration) of a negative outcome highlights a second key point--the parties involved in the RBCA. The article seems to assume that most RBCAs would occur with the health care provider, but providers often have little control and inadequate scale to participate. Prevention /amelioration of negative outcomes is the general province of casualty insurance, so perhaps the appropriate place for some RBCAs is between the patient and the insurance company, not the health care provider. Life and disability insurance providers already account for health behaviors such as smoking and risky hobbies, and they have enough scale, so perhaps it would be possible for them to cover the preventative and acute care related to death or disability from certain diseases. For example, a policy could be written against death or disability from a heart attack, and the insurance company would provide prevention advice/care plus acute and rehab care after a heart attack to reduce their payout.
Finally, most of the examples provided in the article are not outcomes based, but are process based measures. The complexity of medicine and patients makes it tough to find quantifiable outcome measures, but yielding to the temptation of substituting process-based measures will, in my opinion, raise costs while failing to improve the quality of care. There are actually very few process-related measures that apply to 99+% of the population (vaccines are probably one). Most have multiple exceptions and unknowns/changing research findings (e.g., aspirin after heart attack is not recommended if you have high blood pressure, history of stroke in the family, cancer, etc.), which can decouple the RBCA from a significant percentage of cases and create decision tree nightmare. A simplified system can create incentives to treat (or not) totally unsupported by science.
Posted by: SteveSC at Jul 16, 2007 11:46:39 AM
Yes, great blog post title!
Posted by: Daniel Klein at Jul 16, 2007 11:55:36 AM
Pay-for-performance is being studied at several levels in the US healthcare system.
There are difficulties:
1) determining how to measure quality (in less dramatic cases)
2) how to control for patient non-compliance
Posted by: save_the_rustbelt at Jul 16, 2007 4:42:36 PM
Pay-for-performance is being studied at several levels in the US healthcare system.
There are difficulties:
1) determining how to measure quality (in less dramatic cases)
2) how to control for patient non-compliance
Posted by: save_the_rustbelt at Jul 16, 2007 4:43:11 PM
Pay-for-performance is being studied at several levels in the US healthcare system.
There are difficulties:
1) determining how to measure quality (in less dramatic cases)
2) how to control for patient non-compliance
Posted by: save_the_rustbelt at Jul 16, 2007 4:43:16 PM
Thanks to Alex for remembering the article David and I wrote. Being less clever than Alex (great title, really!!), we lost the battle over the acronym (P4P beat out RBCA) but we seem to be making headway in the war. To the comments above pointing out difficulties with and limitations on the use of outcome-based compensation, the appropriate response is that there are few panaceas, and outcome-based payment arrangements are not one of them. We only argued that their use should be permitted and that they have untapped potential to improve the delivery of health care in some areas. I continue to think that’s right. Not all clients pay lawyers on contingency, but some do and there is no reason to prevent clients who like contingent fees from using them.
Posted by: Charles Silver at Jul 16, 2007 11:08:10 PM
I published a related proposal in 1994: Buy Health Not Health Care. Quality is easier to measure when you aggregate treatments, so I proposed paying your health plan based on your overall health outcomes. I don't think one needs to control for patient compliance.
Posted by: Robin Hanson at Jul 17, 2007 9:48:53 AM
Over a large number of patients you don't need to control for patient compliance in part because part of the skill of a treating physician is choosing treatments that the patients will follow.
You do, however, have to control for case difficulty.
-dk
Posted by: Dick King at Jul 17, 2007 2:29:14 PM
Hi Allow me to offer my heartiest wishes. I would like to know more about this I'm an chinese writer who love to read international literature.feel free to contact by sdggc and bjseek-无缝管|无缝钢管|数据恢复|RAID数据恢复
Posted by: 无缝钢管 at Dec 13, 2007 3:05:14 AM
Posted by: 鑽石 at Apr 2, 2008 9:16:36 PM
Why don’t people take their medications? Because they cost too much! I’ve seen ads on TV for Caduet. It has two ingredients. One is Amlodipine and the other is Atorvastatin. With my RxDrugCard I can get 30 tablets of Amlodipine for $9 and 30 tablets of Simvastatin for $9. I’ll bet they are charging more than $18 for this new drug! Don’t pressure your doctor into giving you something just because it’s new. Do your homework. Find a drug card like I did at www.rxdrugcard.com. I think that RxDrugCard.com is the best drug card available for prescription discounts.
Posted by: Lily at Aug 19, 2008 3:04:14 PM
