Network News

X My Profile
View More Activity

Are Doctors Like Farmers?

A doctor commenting at the Wall Street Journal has an idea I hadn't considered:

Pay for performance in medicine will be met with over diagnosis and over coding. Thus, pneumonia will be diagnosed more frequently than it would otherwise, resulting in skewed statistics. If Obama really wants to control healthcare costs, he should pay orthopedic surgeons like me NOT to operate, like crop subsidies.

To be fair, there actually is a way to do this: It's called capitation. Unlike fee-for-service, where doctors are paid for every treatment they conduct, a capitation system gives doctors a lump sum for every patient enrolled at their practice. $15,000, say. If the patient only costs $12,000, then the doctor makes money. In some forms of capitation, if the patient costs more, the doctor loses money, while in other forms, the doctor doesn't lose any money if the patient costs more.

This reverses the incentives. In our system, the money encourages doctors to overtreat. In capitation, the money encourages doctors to undertreat. Britain uses a form of capitation and has had good results in recent years pairing it with pay-for-performance bonuses (where doctors get more money for good patient outcomes) that help protect against undertreatment. I'm actually a big fan of this model, but I rather doubt it will be adopted here.

By Ezra Klein  |  July 9, 2009; 11:30 AM ET
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: The Dangers of Bad Financial Products
Next: Live Chatting


For more thoughts on capitation (all the coolest payment reform kids are doin' it), check out Charlie Baker:

Posted by: AronB | July 9, 2009 11:36 AM | Report abuse

Bah! It won’t result in over diagnosis or over coding. Pay for performance is based on establishing a denominator using risk factors, and a numerator using services performed, e.g., sexually active women over 21 getting paps every 3 yrs/ sexually active women over 21; diabetics with A1C below 9.5*/diabetics; men over 50 getting prostate exams/men over 50; and so on. The lion’s share of pay for performance metrics are about proper treatment of chronic care and preventive medicine. You’ve got a few that are about acute service (post MI beta-blocker prescribing comes to mind, and, incidentally, was met by physicians so quickly after it became an indicator that there’s talk of removing it as an indicator), but most really aren’t.

And, re: capitation, it has it's benefits, but it is a disincentive for data-sharing. If docs are getting paid by the month, it's just extra work for them to submit all the data the various entities want to track quality.

*Get 10 endocrinologists in a room and you’ll get 10 different values on this.

Posted by: ThomasEN | July 9, 2009 12:09 PM | Report abuse

Capitation? Been there, done that, bought the T shirt: capitation didn't work, patients hated it, doctors hated it. Apparently you don't remember the 1990s. I very much doubt patients and doctors would like capitation any better were it imposed by the federal government as opposed to HMOs.

What's that old hackneyed Santayana quote? "Those who do not remember the past are doomed to repeat it."

Posted by: Policywonk14 | July 9, 2009 12:38 PM | Report abuse

Good post. When we speak of pay-for-performance we don't always think of all the situations and aspects. Diagnoses, drug costs, etc.

But all of these aspects can be handled quite well!

I've been working on this for weeks, and there is a very good way to pay doctors, that is general and handles complexities of all sorts.

This structure has been refined after discussion with a retired surgeon.

Here is a brief template on how to set up good outcome incentives that includes various complexities like complex diagnoses, indeterminate outcomes and typical situations:

Posted by: HalHorvath | July 9, 2009 12:55 PM | Report abuse

Did you miss the 1990s? Maybe you've never been under capitation. Capitation creates incentives for doctors to encourage sick patients to switch elsewhere or to not come in for problems. Obviously, there's a limit to this (you don't want them dying and suing you), but I believe it's pretty strong until you get to life-threatening issues.

My doc's office wouldn't pick up the phone line for several days in a row, and they'd of course never call you back. The message said go to an ER if it's an emergency. Talking to a doctor without coming in to pay the co-pay? Forget about it. Long lines? Too bad. Have arthritis? Deal with the pain. Have skin problems? Buy cream.

I believe capitation leads to under-diagnosing problems and not dealing with those which won't cause death.

Posted by: GrandArch | July 9, 2009 1:06 PM | Report abuse

We did have a fair amount of capitation in the 90s and it got caught in the general anti-HMO anti-gatekeeper backlash. I do not think patients disliked capitation per fact, most patients do not know how their docs are paid.

A word in favor of capitation---the systems we typically hold up as truly working (Kaiser, Group Health, Geisinger, Scott&White) have elements of capitation.

Paring capitation with pay for performance would strike a nice balance.

Posted by: scott1959 | July 9, 2009 1:09 PM | Report abuse

GrandArch, your example is an instance handled specifically in the structure of pay-for-outcome I lay out in that link above.

Posted by: HalHorvath | July 9, 2009 1:22 PM | Report abuse

The thing about capitation is that it turns individual medical practices into insurance companies: the practice must take on the risk that their individual group of patients will need a significantly above-average amount of care. For a large system like Kaiser, which insures and cares for millions of patients and hires thousands of doctors, this is not a problem -- the pool is large enough that the costs will be very close to the average. But for a smaller practice with only a few doctors and a few hundred or a few thousand patients, the risk is much larger: to benefit, they would need not only to keep their costs low, but also to be "lucky" with respect to the needs of their patients.

Posted by: alexrp | July 9, 2009 2:15 PM | Report abuse

Usual compliant - Ezra mentions England, but the rest of the commenters seem to forget that other countries must have various solutions and data on them. Why don't we examine what they do more carefully? After all at the end of the day they get better health care than we do at less than half the cost.

Posted by: lensch | July 9, 2009 2:22 PM | Report abuse

I'm not sure you can really declare capitation a failure yet - I don't think its really been tried in much of the country. I am very familiar with some of the larger staff model, non-profit, member-run plans. Maybe "non-profit" and "member run" serve some of the same functions as pay-for-performance but these are the rare plans that create an incentive to develop and implement even simple cost saving programs. Personally, I'm much more comfrtable with a doctor who will explain why I don't need the antiboitic (to say nothing of the surgery) than one who gets paid to go overboard with the procedures.

Any national plan will have my support if they can persuade me that it will for example reward the following:

A post-hospitalization nursing or follow-up program that reduces re-hospitalization rates.

A dedicated RN/LPN to providing training and followup for diabetic patients which reduce complications and doctor's visits.

email or telephone consults with doctors or nurses that reduce office visits.

These are the types of plans that many providers will not implement now because they lose money under the current system if the plan is successful.

Posted by: metis1 | July 9, 2009 2:36 PM | Report abuse

Capitation works, but it has to be risk adjusted. If all patients were comparably healthy, then straight capitation would work fine. But that's not reality. With risk-adjusted capitation doctors receive a higher premium for a sicker patient. So, the 40-year-old, non-smoking marathoner may warrant a $12,000 annual premium for the doctor. But the 60-year-old, obese, smoking, diabetic warrants an $18,000 annual premium.

Without this doctors have incentives to cherry-pick healthy patients (as insurance companies already do, which would further exacerbate the plight of those in poor health) because it would wreck their profitability model. Additionally, not only does risk adjustment incentivize doctors to only provide care that is necessary (as suggested by the subsidy premise), but it also incentivizes wellness care. Under the fee-for-service model doctors have no financial incentive to spend time with their patients consulting about lifestyle and other factors that could substantially improve their health.

As an example, consider the diabetic with poor circulation who is at risk for a future amputation. A FFS doc has a financial incentive to examine him, prescribe is drugs, and send him on his way so he can get to the next patient. A RAC doc has an incentive to spend some time with him, look at his feet, discuss his lifestyle, and keep him healthy (read: keep his health costs low). The longer that patient is alive and healthy the longer he is a revenue stream for his doctor.

Posted by: gpatterson1 | July 9, 2009 2:39 PM | Report abuse

Just noticed your final comment about capitation not being adopted here... Medicare already is risk-adjusted and capitated!

Posted by: gpatterson1 | July 9, 2009 2:41 PM | Report abuse

alexrp explained the problem with capitation.

That being said, I still think that it's a grand idea for primary care. You get paid $X per month for each patient registered with you. They pay a copay when they come in (it does cut down on the stupid visits). No billing department, no contracting department, hugely reduced overhead for PCPs. If you provide good service, then you have patients beating down your doors and you make a lot of money. If you provide crappy service, then the patients go elsewhere and the doctor starves. If a patient needs a referral to a specialist, the PCP shouldn't be on the hook for the entire cost, but a small payment deducted from the monthly capitation, keeps the PCP from referring everything out.

When capitation was at its zenith in southern California in the 1990s, 50% of doctors groups failed and were replaced with megagroups. It's really the giant beaurocracy that people hate when they hate "HmOs".

Anyway, that's what I would do, if I were dictator. Single payer with capitated primary care.

Posted by: J_Bean | July 9, 2009 4:27 PM | Report abuse

Or you could make it easier for people to get a second opinion. If you only have one option for medical advice and that option tells you that you need surgery then you'll most likely get the surgery.

It seems that insurance companies would actually have an incentive to encourage their clients to seek second opinions if it could potentially reduce their payments for hypothetically unneeded surgery.

Why do you always prefer options that increase centralized control?

Posted by: fallsmeadjc | July 9, 2009 4:33 PM | Report abuse

alexrp....capitation with risk adjustment and outlier clauses will take care of your concern here.

gpatterson1...Medicare is a fee for service system. No capitation or risk adjustment.

Posted by: scott1959 | July 9, 2009 7:11 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company