It'll take more than Shannon Brownlee's bloviations to make this doctor's head explode

Mike the Mad Biologist posts a link to and excerpt from an article that he seems to think will make "M.D.'s heads explode."

It didn't. At least, not in the case of this M.D.

Basically, it's about physician reimbursement, a topic guaranteed provoke controversy, divided between those who think doctors are already overpaid (most non-physicians) and those who do not (most doctors). The article is by Shannon Brownlee and makes a proposal that is breathtakingly naive and poorly thought out:

...we don't end up saving any money by tightening reimbursements. But we do end up pissing off doctors, who don't really want to have to run around doing more procedures and seeing more patients just to maintain the same income. It also means we patients can expect to be given a lot more unnecessary procedures, because when doctors do more, they don't necessarily do more of what we really need. In the 1990s, managed care trimmed physician reimbursements and an avalanche of unnecessary procedures and blood tests and CT scans was the result.

All of which is just one more reason why fee-for-service has got to go. It's a broken payment system, and it simply encourages bad quality care. Doctors need to be put on salaries.

Alan Sager, at Boston University, suggests that we take the portion of our national health care bill that already goes toward physician reimbursement -- about $500 billion -- and say to doctors, in effect, you can keep the money, but you have to take it in the form of a salary. Surgeons would no longer be paid separately for each surgery, and primary care physicians would no longer get a separate fee for each office visit.

We might want to redistribute the money a bit, so primary care doctors make a little more and the super-specialists make a little less. The main idea here is to recognize the fact that we can't save money by squeezing doctors, and what we ought to be doing is removing the financial incentives for giving patients care they don't need.

Actually, being an academic surgeon, I already work for a straight salary. I could make more in private practice, but the differential has been decreasing, thanks to my salary gradually increasing and the take home pay of private physicians remaining stagnant.

As for this "solution," it neglects a number of things, not the least of which are for-profit health insurance companies. I'm also very skeptical of the claim that there was an "an avalanche of unnecessary procedures and blood tests and CT scans" due to reimbursement cutbacks in the 1990s. For one thing, ordering blood tests and CT scans do not add to a physician's income. Even if I were in private practice, ordering such tests would not make me a red cent, although it would make money for the lab and for the radiologists who do the tests. Given that there are laws preventing physicians from owning testing companies to which they refer patients, this is not really an incentive. True, a small proportion of such tests might uncover abnormalities that need treatment, but that's a really long run for a short slide, given the low yield. I'd have to look at lots and lots of tests to find one patient who needs a procedure.

In other words, although the idea of salaried physicians has some merit, this Brownlee article is full of dubious assumptions on a number of levels.

As for the concept of a salaried physician, which, being one myself, I don't have an intrinsic dislike of the concept. It works just fine for me at the moment. However, Brownlee's solution leaves out the fact that a significant fraction of reimbursement does not come from Medicare and Medicaid, but from private insurers. The government could try to become the employer of all physicians by paying physicians on salary, but that leaves out a huge area of reimbursement. And if this proposal involves taking that private insurance and including it in the pot that provides physician salaries, then that is a de facto nationalization of the physician work force. If that's what Brownlee believes to be the solution to the problem, fine, but she should come right out and say it, then defend the proposition for what it really is.

Here's the real problem though. Brownlee's concept of making all physicians take a straight salary, presumably dictated by the government would also be a bureaucratic nightmare to an extent that would not necessarily any better than the present system and might very well be as bad or worse than our current hodge-podge system of piecemeal reimbursement. As a commenter named jon said:

Doctors presently get money from government healthcare, corporate healthcare, and private payers (and combinations of all three.) How would they merge?

Would a doctor that sees 90% private persons paying cash and 10% Medicare/Medicaid get 10% of the government maximum salary? What about a guy who sees 10% private, 30% private/corporate, 30% government healthcare and 30% government/private patients but only fully treats the private patrons? I see the doctors not having to let any of their office staff go. Right now, handling the billing is more than a full-time job or two for most doctors' offices. How this can be fixed will have to be part of the solution rather than an addition to the problem if this proposal is to have any chance of flying.

Shannon's response to this comment was not particularly enlightening. In fact, it was in essence a non-answer:

Richard Riley gets the prize for coming to precisely the next question -- who pays, and who does the physician work for?

First, let's remember that there are lots of doctors on salary -- at places like the Mayo Clinic, and Kaiser Permanente, and Group Health Cooperative of Puget Sound, the Veterans Health Administration, and Intermountain Healthcare, some of the best organized medical practices around. The key word here is organized, but I'll leave that part for another post.

These systems are either prepaid, integrated Health Maintenance Organizations (to be distinguished from the managed care we all learned to hate in the '90s). Or, in the case of Intermountain Healthcare, an integrated hospital system that hires its physicians. Or a government-run system with a budget, a la the VA.

They are also demonstrably better at delivering high quality care than the rest of the fee for service world.

So who would salaried doctors work for? Either an HMO, as is the case in Group Health and Kaiser; the government, as in the case of the VHA; or a hospital system, as in Intermountain Healthcare.

Yeah, yeah, yeah, I know, everybody is blowing a gasket right now, saying I'm crazy if I think Americans are going to go for either a government-based system like the VA, or an HMO. All I can say is, look at the data. These systems all do a better job of delivering care that people need, not giving them stuff they don't need, and doing it all for less money.

Oh, and the doctors who work in them are generally more satisfied with their jobs.

Who would pay? Well, employers and individuals pay HMOs directly. All of us who pay taxes help support the VA system. I don't know how Intermountain gets paid.

None of this addresses the main problem: How would each physician's pay be determined, particularly the proportion that would come from each source? Who would "organize" the distribution of payments? Which HMOs would physicians work for? How would they be forced to work for such HMOs? (And, make no mistake about it, this would require a carrot and a huge stick to accomplish.) What evidence does she have that her system would be any less complex than the present system or that it would prevent more procedures? Clearly, Brownlee hasn't thought this through very well, if this post is any indication.

In fact, her second post on the topic is also full of assumptions that are not necessarily related to reality. For example:

One poster stated: "You can't just say that you're going to cut out the 5% of procedures that are harmful." OK, but instead of focusing on harmful care, let's look at unnecessary care (which, it turns out, is also harmful). The amount of unnecessary care delivered in this country isn't 5 percent. It isn't even 10 percent. It's closer to 20-30 percent. That means we're wasting between $400 billion and $700 billion on care that isn't doing us any good.

How do we know this? The best work has come out of research at Dartmouth that has uncovered wide variations in how much care patients receive in different parts of the country. The Dartmouth researchers have found a two-fold difference in per capita spending on Medicare recipients in different regions.

That difference can't be explained by variations in the price of medical services. Of course prices differ in Biloxi and Boston, but not enough to account for the two-fold variation in spending. It also can't be explained by differences in the prevalence of illness in different parts of the country. (Yes, people are generally sicker in Biloxi than in Seattle, but that can't account for the difference in spending either.)

The only explanation for it is that doctors and hospitals in different regions are giving Medicare patients a lot more -- and a lot more intensive -- care.

No, that's not the only explanation. To say so is, quite frankly, dumb. It may be part of the explanation, but regional variations in care patterns and usage represent a highly complex topic for which a number of explanations have been proposed. She utterly neglects referral patterns, malpractice concerns leading to "defensive medicine" and more testing and referrals to specialists, and even differences in training at different residency training programs that produce the physicians who live in the area. I am not by any means trying to claim that overtreatment is not a problem, and, indeed, the only point that Brownlee gets mostly right is that more specialists do tend to lead to more referrals and higher costs. However, it should be noted that these specialists tend to congregate in urban areas considered desirable, such as New York, Boston, Chicago, etc. Putting physicians on salary would not fix this aspect of the problem; specialists would still tend to congregate in such areas. To use Brownlee's example, there would still be more specialists in Boston than in Biloxi. Because doctors, tending to have a a generous income by and large also tend to like the culture, lifestyle, and amenities that large urban areas provide, likely the only way to truly overcome this problem would be to dictate the geographic distribution of physicians, as, for instance, Canada does.

In other words, command and control.

All in all, although converting physician reimbursement to a salary system is not intrinsically a bad idea, Brownlee oversells its potential benefits and seems not to be aware that it would not necessarily be any better or any less complex than the system it replaces. It also does nothing to explain exactly for whom all these new salaried doctors would be working or how the various funding streams that reimburse physicians for their work would be integrated into a salary system. She seems to assume that it would not be that big a deal to do.


More like this

Yea -- I think my physician is overpaid -- right until I do something stupid at 3 AM and she meets me at her office to treat my whinning butt. Would she do that if she was on salary? Friendship only goes so far.

Great article.

Brownlee also ignored the effects of tax laws governing not-for-profit organizations on utilization of equipment. To maintain that status, not-for-profit health care organizations have to spend their revenues, and frequently do so by ordering additional or new equipment that cannot be supported by the population served (especially if another provider already has that equipment). To justify the purchase, which is also required to satisfy an auditor's questions, the equipment has to be used. So, it is.

There's a dynamic involved. While a provider can legitimately decide that sending a patient across town for a diagnostic procedure isn't worth it, since it will take a while to schedule an appointment and get the results back, by which time the provider may have ruled out by other means the condition that the procedure was to test for, it's a different situation when the procedure can be done a couple of floors away and the delays avoided. However, as ordering the procedure becomes more common, it becomes accepted standard of care. Then, at least as a matter of defensive medicine, the procedure has to be ordered, whether or not the provider, exercising sound professional judgment, thinks it's really necessary.

You sure do like that "long run for a short slide" phrase.

I was in a large clinic. When I started, we were around forty docs, with four general surgeons. When I burned out, we were over 200 docs, and three general surgeons. I saw 1200 more patients per year than one of my partners, and 1000 more than the other. I took less time off, and always worked on my "day off." In part, it was because when docs called with a patient needing to be seen urgently (as opposed to emergently), I was usually the only one willing to stay late, come in early, work over lunch, etc. Compulsive about being on top of things, and about getting patients home from the hospital as efficiently as possible, I made rounds at minimum twice a day, often three. My partners: once. In my scheduled surgery blocks, I could do half again as many cases as they or most anyone else, for various reasons, including (but hardly limited to) the fact that I helped turn the rooms over, always hand-carried lab and other data to the OR the night before...

Some people have a stronger work ethic than others. Still, I can say with certainty that if I'd been making exactly the same as my partners no matter the differences, those differences would have levelled out, if not entirely disappeared. I'm human, as it turns out. The effect? Longer waits, less happy customers. Within two years of my retiring, the clinic went from three to six surgeons. A year later: seven.

This is not unlike the problem Congressmen have. Or guys like Tony Snow who recently quit the White House because the money ($168K) wasn't right.

People with high energy and smarts have attractive alternatives and high expectations. Folks laboring away in lower cost areas with less drive and/or training aren't going to relate. The media is the wrong place to stage this because it's not appealing to the mass of people struggling to pay their bills on $20 an hour.

That said, being on a salary doesn't mean you can't have incentives. An arrangement where you have a salary and a productivity bonus has is the way to go.

"For one thing, ordering blood tests and CT scans do not add to a physician's income. Even if I were in private practice, ordering such tests would not make me a red cent, although it would make money for the lab and for the radiologists who do the tests."
True but it might bring the back for a second visit and thus a second fee.
The trouble with straight salery is it allows the lazy to make as much as the industrious. How about an hourly rate?

The salaried-doctor argument is basically a cross between a red herring and a strawman. If someone thinks putting doctors on a salary will solve the health care problem in the US, he or she is not thinking straight. Can we lay this debate to rest and, if we are so inclined, argue about more relevant things?

Unnecessary procedures are not incentivized by the most prevalent current managed care physician payment systems, physician overwork is. They aren't the same thing.

Fee for service, which is not the most prevalent current system, does incentivize unnecessary procedures, and there were a lot of articles like Shannon Brownlee's written 15-20 years ago when fee for service was more widespread and unnecessary procedures were thought to be more of a problem.

Even 15-20 years ago, few commentators were proposing salaried physicians, since the obvious incentive would be to provide as few services as possible, and people worried whether necessary care could be obtained in timely fashion under such a system.

The answer that was eventually arrived at by many private insurers was capitation, in pure or modified forms one of the most common types of managed care reimbursement today. This essentially pays the physician per patient, so there is no longer an incentive to provide unnecessary procedures. Rather, the incentive is for the physician to treat as many patients as possible by spending as little time as possible with each patient. Both overworked doctors, and patients who spend only 15 minutes of an hour in the doctor's office actually seeing the doctor, should be nodding in recognition at what this reimbursement system has wrought.

Each reimbursement type has its associated problems, and it's rather surprising that anyone is advocating one method as "the solution."


You're right in terms of primary care and medicine. Surgeons, however, are still primarily paid by the procedure, as are interventional radiologists and cardiologists and pretty much any procedure-driven specialties.

My old boyfriend, an anesthesiologist who works on a liver transplant team at a major NYC hospital, told me he makes $30/hr for medicare patients. Far less than what the plumber made per hour to come unclog his toilet. A needed skill, yes, but I think it takes a little less sacrifice in the training stages.

I called the method of paying doctors a red herring because it's only a small part of the cost of health care. I recently had rotator cuff surgery. My insurance paid about $90 out of a cost of about $1000 for the MRI. I have not yet received the surgeon's bill, but I expect it to be only around 20 percent of the cost of the surgery. The surgery center bill appears to have been around $27000 to $28000 (my insurance paid $25,000 and I will end up paying more than $2,000). Even with insurance (bad insurance, since I am self employed and have only what a reasonable person might pay for), my out-of-pocket expense is likely to be at least $6000 for this episode, all caused by a fall from the second step of a ladder. There are plenty of working people around for whom this would be a life-changing experience if they did not have insurance, and a major roadblock in life even if they did have insurance like mine. Putting doctors on a salary would do essentially nothing to solve that problem.

Arguing about doctors' pay is a distraction, but doctors' incomes are a very tempting target. My brother in law is an orthopedic surgeon. He has a very nice little horse farm, an airplane and a second home in Ireland. By most reasonable standards he would be considered wealthy, partly from medical income and partly from wise investment of that income. What is a working stiff (like carpenter, painter, other crafts) to think when he sees that and knows that paying for surgery that will allow him to keep on working at a low-paying job will put him in debt for years if he can even afford it? This episode will not break me, fortunately, because my income is also relatively high. But I still can't see how anyone can possibly argue that the current system of health care in the US is not broken, and badly.

Yeah, doctors' salaries are a small part of the problem, and is the least urgent to deal with.

Insurance is the major problem. What was a good idea is now a monstous bureaucracy with an ever changing system of codes, rules, and subterfuge in the name of making money.

I have nothing against making money; I do have a problem with people and organizations making money in an unethical manner. It may be legal, but it is a monkey on our backs.

While I haven't a clear idea how this problem can be solved, I do think it needs to be done in such a manner that: doctors within such a system maintain their autonomy and ability to treat their patients as they see fit without unreasonable restraint or oversight by people who know less about medicine than a nurse's aide; patients also maintain their autonomy in that they have a choice as to what procedures and treatment they want, i.e. the right to refuse treatment or to make bad choices; and finally, if insurance under any name survives, to cover those procedures deemed the best choice for health or relief based on EBD. Or a physician can choose to go private pay.

Now how the hell that would work is still a mystery to me.

From: a very poor philosopher indeed.

As an economist I always end up smacking my head against my desk when I hear people suggest price (or wage) fixing as a way to solve market problems. Shannon Brownlee: why do you hate my skull?

The price mechanism is the most subtle and power informational system you will find anywhere. Tampering with it is extremely dangerous and I guarantee you the result would be severe adverse consequences. Geographic imbalance of specialists would only be the beginning.

Rjaye: It seems to me that the problem is employer-funded insurance, as it makes it hard for dissatisfied insurees to change insurer. With a captive market, insureres have no incentives to behave, provided they aren't so egregious that they alienate the employer. There are other barriers to changing insurer (like pre-existing conditions), but there are ways of addressing these problems.

Orac wrote: "Surgeons, however, are still primarily paid by the procedure, as are interventional radiologists and cardiologists and pretty much any procedure-driven specialties."

Sure, though even re surgical procedures Medicare has for many years structured reimbursement to minimize multiple procedures in a variety of more or less subtle ways. For example, for outpatient procedures, docs only receive full Medicare payment for one procedure per patient per day. Any associated procedures for the same patient on that day are paid at a substantially discounted rate.

Another way Medicare attempts to remove incentives for multiple procedures is the Stark rules, which prohibit a physician from referring patients for, e.g., lab work to labs in which the physician has a financial interest. Thus, while a surgeon may be paid per procedure for surgery, he/she may not be paid for associated procedures like lab tests.

I talked to my wife about the cost of my rotator cuff surgery, and I had forgotten about paying the surgery center $4400 up front, so the surgery center alone will be about $32,000. Our cost is really approaching $10,000. And that's with insurance. Don't tell me the health care system ain't broke.

It seems to me that the problem is employer-funded insurance, as it makes it hard for dissatisfied insurees to change insurer. With a captive market, insureres have no incentives to behave, provided they aren't so egregious that they alienate the employer.

Insurance is the main problem for another reason as well -- it distorts the economic relationship between provider and consumer of health care. If a patient has to pay for expensive care himself, he is more likely to shop for price than a person who is not paying himself. When someone else is paying, there is little incentive to minimize usage or control prices, and demand becomes essentially infinite. As demand for a service (health care, in this case) increases, while supply is limited, the price necessarily increases, and this is indeed what we see.I think the various plans that restore the linkage between those using health care and those paying for it (for example, health savings accounts) have the best chance of restoring sanity to our broken health care system.

We are actually seeing more and more doctors moving onto salary, partly because of the rise of HMOs, but also because it is harder and harder for a single doctor to effectively run an office profitably.

As far as a private practice doctor is concerned, health care is an oligopsomy. Doctors' services are bought by a small number of insurance companies, local, state and federal government. A doctor in private practice has NO pricing power. When the buyers cut reimbursements, it's a matter of take it or leave the plan and lose the patients.

Larger organizations, clinics, hospitals and MD service resellers have some pricing power. They can afford to lobby the government, and they can negotiate with the insurers if they are big enough. They can also play games and restructure their business to keep up the cash flow, hence the added tests, added procedures and more serious diagnosis codes. They can benefit nicely from such maneuvering. They usually hire doctors at fixed salaries, but there is usually a required level of billing, as at a law firm.

From what I have seen, newly qualified doctors are increasingly reluctant to strike out on their own, and are less likely to do so in five or ten years than doctors twenty years ago. Opening a medical office requires a fair bit of business acumen, and a fair bit of time spent managing the business. A doctor has to lease or buy an office, remodel it, buy equipment, hire an office manager, a receptionist, a team of insurance and medical support specialists and so on. Not every doctor wants to be a small business person, especially one with no pricing power. It's not as if a sterling reputation and an amazing cure rate will let a doctor earn a dollar more than any other doctor reporting the same diagnostic code.

Yes, surgeons do bill by the procedure, but this is within the framework of the hospital. If a surgeon bills $2,000 for an operation, odds are the hospital is billing $10,000. A surgeon might be able to justify a private practice as a means of feeding their surgical practice, but even here the economics are marginal. There are salaried positions with admitting privileges, and this is increasingly common.

Whether we move from a small number of payers to a single payer or not, these pressures will persist, and more and more doctors will wind up on salary. The down side is that it will be harder to make a fortune as a doctor, unless one specializes in diseases of the rich. The up side is that doctors will have lives, and if they are smart enough to organize, as many have done, they will be able to live rather comfortably and focus on providing good medical care.

James wrote:
>>Rjaye: It seems to me that the problem is employer-funded insurance, as it makes it hard for dissatisfied insurees to change insurer. With a captive market, insureres have no incentives to behave, provided they aren't so egregious that they alienate the employer. There are other barriers to changing insurer (like pre-existing conditions), but there are ways of addressing these problems.<<<

James, that's an excellent point, and definitely needs addressing. It also limits competition between insurers.

Stevge: The separation of payment and treatment is a problem, but the same phenomenon occurs in other insurance markets, so its probably not a fatal problem.

Rjaye: For markets tob e effecieint they need the "discipline of the market", the creative destruction of competition. That's why I think a singler-payer system is a bad idea. That would be even less competative than what you have now.

Because doctors, tending to have a a generous income by and large also tend to like the culture, lifestyle, and amenities that large urban areas provide, likely the only way to truly overcome this problem would be to dictate the geographic distribution of physicians, as, for instance, Canada does.

Orac, where'd you get this idea? Doctors in Canada are free to practice anywhere they want. Some provinces (i.e. Alberta) provide generous bonuses to people setting up shop in rural areas, but most doctors establish themselves in the urban area of their choice. (Quebec does this ridiculous thing where they cut your billing by 30% if you work in Montreal, but Quebec is... different.)