Overtreatment

Shannon Brownlee's guest-blogging at The Washington Monthly about the aspect of the health care issue I don't see any politicians wanting to tackle -- the fact that doctors frequently overtreat patients in ways that are sometimes directly harmful and even when not harmful per se, contribute to a terrible maldistribution of health care resources. That's not to say that America has "too much health care," but rather that at the same time as many Americans have too little health care other Americans are, in fact, getting too much. Doctors are, in essence, prescribing all the treatment that will get paid for -- which means too much treatment for people with a large ability to pay, and too little for people with little ability to pay.

Brownlee's alternative is to turn doctors into salaried employees charged with doing the job of keeping people healthy, rather than into fee-for-service professionals whose level of compensation depends on how much treatment they prescribe. That seems appealing to me, but it's considerably more radical than anything being contemplated in the political system right now.

Comments

Several years ago I read about a computerized expert system diagnostic tool that had a better accuracy rate than humans. There are probably exotica of diagnosis that require human intervention, but there are also refinements to expert system tools that can be implemented regularly to expand the realms where we don't need highly paid people. It's easy to imagine a medical facility where most patients are diagnosed with by a PA or RN using a computer aided tool. The expert system would then determine the lab work and the PA/RN would only over-ride the recommendations where there's obvious error. Costs and procedures could be monitored and standardized. Most medical care isn't brain surgery. After all, currently half of all doctors finished in the bottom half of their classes.

(When I need care, my first line is a PA. I like and trust the PA I go to.)

Posted by Jeffrey Davis | September 4, 2007 12:14 PM

Removing the fee-for-service incentive might help reduce overtreatment, but I think you'd have to immunize doctors from lawsuits too. One reason to pile on treatment is to make sure you've done everything so that a patient (or his lawyer) can't say you neglected this or that standard of care. Of course, that doesn't explain the neurosurgeon here in Rhode Island who operated on both sides of a patient's brain because he first started operating on the healthy side....

Posted by Bob | September 4, 2007 12:17 PM

Removing the fee-for-service incentive might help reduce overtreatment, but I think you'd have to immunize doctors from lawsuits too. One reason to pile on treatment is to make sure you've done everything so that a patient (or his lawyer) can't say you neglected this or that standard of care. Of course, that doesn't explain the neurosurgeon here in Rhode Island who operated on both sides of a patient's brain because he first started operating on the healthy side....

Posted by Bob | September 4, 2007 12:18 PM

The over-treatment issue is real, yet can lead to some false conclusions. Within academic medicine there has been for some time a movement for "Evidence-Based Medicine" that requires a high standard of proof of efficacy with real outcome data. Not coincidentally, the modern version of this started in the United Kingdom and Canada. The slow move to clinical practice guidelines and standards, and implementation of CQI in this country has been helping where they are done. Where they are done tends to be in government controlled settings such as university-teaching hospitals, the VAMC, HRSA's community health centers, and the like.

But private practitioners without some external oversite can do anything they want, so long as somebody will pay for it.

Needless to say, I think that single payer, expanded and improved Medicare for all would help this situation, by reimbursing only for proven treatments.

It is not all a matter of defensive medicine, although the legal incentive to treat vs. not treat does bias towards treatment (it's only a viral cold, but lets throw antibiotics at it anyway just in case).

There is also a societal presumption and expectation, by both patients and doctors to treat.

And there is very little reimbursement for just seeing a patient, a little more reimbursement for medical treatment, and a whole lot more payment for doing procedures... even if they are unproven.

The problem is the same therapeutic nihlism that is sometimes expressed in articles like these (and going back to McKussick), quickly gets used as an excuse to NOT provide coverage to the poor, etc. It winds up being used by some as an argument that health care generally is a luxury and not a necessity. Beware.

Posted by DrSteveB | September 4, 2007 12:24 PM

One aspect of this problem not addressed here or in the original postings (including KevinDrum et al) is that Dr's and Hospital Administrators feel they are doing what the public wants. And they try real hard to do it. For many MDs, the patient's requests for tests might make little sense ('Do you think I ought to have a CATScan?'). But they don't feel they have a lot of choice. First, they can't get sued or reported to the state Board for doing what the Patient requests. And second, they get the reimbursment. Third, their colleague (the radiologist who might read the SCAN) also benefits which is nice in a professional way.

There are lots of good reasons to be an informed consumer of Medical/Hospital services. But lots of Dr's would tell you it's gone too far.

(Disclosure: I'm an RN in a family full of nurses and docs)

Posted by JohnMcC | September 4, 2007 12:47 PM

I'm a doctor who thinks that our current system combines the compassion of capitalism with the efficiency of socialism. Obviously, we need to make changes.
However, I also think that any real existing solution to a problem will always look worse than some theoretical ideal.
I suggest that we take advantage of the fact that we are the last developed country to adopt a natioal health care program. Let's begin by studying the experience of other nations. Let's do it without any preconceptions. Let's do it without an agenda that looks for an electoral advantage for one party or the other.
I know that I'm a dreamer, but I'm not the only one.

Posted by Slugger | September 4, 2007 12:50 PM

I'm with Slugger. One of the primary benefits a review of single national health care systems is that they reduce malpractice as a legal problem. Since everyone has to be treated anyway, a large portion any malpractice claim disappears. Overtreatment is thereby reduced. This is about number 23 on reasons why national health is the way to to go.

Posted by Peter Driscoll | September 4, 2007 1:05 PM

There are ways of paying doctors other than salary or piecework. One way is capitation -- a doctor is paid by the third party insurer for the number of people he treats annually, regardless of the procedure. Put a cap on capitation and give a bonus for wellness outcomes and you can start to get somewhere useful.

Posted by Francis | September 4, 2007 1:21 PM

Doctors do not necessarily benefit financially from prescribing extra treatments or tests. Outside the office visit environment (where the amount of additional expense this creates is apt to be fairly small) they do not benefit at all. Instead pharmacy companies or hospitals capture the extra revenue. So turning doctors into salaried employees is not a solution. Over-treatment probably does not have economic causes, at least not direct one. Overtesting may be a case of "CYA" out of fear of lawsuits, but I suspect that over-treating is more generally motivated by physicians who find themselves frustrated when facing difficult or terminal illnesses and decide to throw everything at the problem they can, at least when they know the patient can afford it.

Posted by JonF | September 4, 2007 1:30 PM

turn doctors into salaried employees charged with doing the job of keeping people healthy, rather than into fee-for-service professionals whose level of compensation depends on how much treatment they prescribe.

The Mayo Clinic already operates this way. Would be interesting to see if it could be applied to health care providers other than those at the most elite hospitals.

Posted by Matt F | September 4, 2007 1:34 PM

Bush will pronounce "nuclear" properly before doctors convert to a salaried basis of compensation.

Posted by steve duncan | September 4, 2007 1:46 PM

I think it's high time we had socialized punditry in this country. My proposal is for a huge federal bureaucracy that dictates that pundits and political writers get paid on salary rather than by piece. This new agency, called, say, the Bureau of Punditry Affairs would set a given pundit's salary based on the years of experience and his topical specialty. The more general the analysis, the lower the salary. Obviously pundits will make less money than they do now, but it will greatly assist poor people who have limited access to valuable and timely political analysis.

Posted by Homer | September 4, 2007 2:16 PM

Ever take your car to the garage and then find out they did things that didn't need to be done? Right.
The people that fix cars are paid on a "fee for service" basis, so the more they do, the more they get paid. Hey, isn't that the way we pay doctors?

Posted by kafka | September 4, 2007 3:22 PM

Jeffrey Davis,

The WSJ's Holman Jenkins had a column somewhat along those lines a few years ago, when he suggested that the relationship between physicians and automated diagnostic programs should be akin to that of airline pilots and autopilot. He thought this could dramatically reduce the error rates in medical treatment.

Posted by Fred | September 4, 2007 3:36 PM

Canada has a single-payer system, but pays most doctors a fee-for-service. The malpractice lawsuit worries are not as big as in the U.S. (though they exist). Instead we have docs forced to run lots of patients through their practices, with lots of services charged, in order to cover their overhead etc. There is very little time to make a personal connection with your patient and to promote wellness. My young doctor friends are all fleeing family practice because of these sorts of issues.

There are experiments in several communities in Canada (e.g. Sault Ste-Marie in Ontario) where salaries and capitation are being tried, and they seem to be working well.

Of course, the cost of prescription drugs, inequalities in serving urban vs. rural communities (and the dreadful health care received by aboriginal people), resistance to new management processes, etc. are all big issues too...

Posted by GC | September 4, 2007 3:41 PM

the fact that doctors frequently overtreat patients

The last time the media, etc., made a big deal about this, in the end all that it did was give insurance companies cover to start denying claims left and right.

Brownlee's alternative is to turn doctors into salaried employees charged with doing the job of keeping people healthy, rather than into fee-for-service professionals whose level of compensation depends on how much treatment they prescribe. That seems appealing to me, but it's considerably more radical than anything being contemplated in the political system right now.

This is, in theory, what HMOs are supposed to do. My dad is a health care provider of a sort working for an HMO. He doesn't get paid on a per procedure/visit basis. He's scheduled to see a certain number of patients per day, and they pay him a salary.

Back in the day, this was an excellent deal for a medical professional: you didn't have to worry about the business end of medical practice, you had full secretarial support, you even had a schedule all made up for you with ample time alloted for lunch, to spend extra time with patients should the need arise, etc.

Of course, now, since the emphasis is on profit, profit, profit, the HMO schedules more and more patients in a day, dumps a lot of the paperwork onto the medical providers as they loose secretarial staff through attrition (not replacing those who have retired), etc.

But the theory of an HMO -- if certainly not the practice -- is hardly radical and it is exactly what MY seems to be getting at here and is actually exactly how health care should work, even if politically it'll be difficult to impliment considering how badly the practice of for-profit HMO's has discredited what seems to be the best idea around. The problem, of course, fundamentally, is greed rather than anything else, ain't it?

*

And just a general comment (I say this even though I am probably less than a decade older than MY and his peers): MY, et al., y'all are very smart and have tons of good ideas. But it would behoove you yung-ons to learn a bit of history and have some perspective when you make comments about what's a radical idea and whether or not we should be concerned about over-treatment. These things all have histories (as does having health care as a benefit) that y'all better know before you start pushing things that have boomeranged in the past.

Posted by DAS | September 4, 2007 3:43 PM

The Sutton Principle

It really isn't even a somewhat new insight that too much medical intervention is as undesirable as too little. Interventions are limited to licensed providers only insofar as they are harmful. You don't need a license to eat right, exercise more, lose weight and generally follow your mother's sound advice, because you can't hurt yourself with these interventions, while it is precisely these interventions that can do the most good. You need a license to do bone marrow transplants, because this intervention is more deadly than all but the worst prognosis cancers, so you better be right that you have exactly the cancer that justifies a specific intervention, and then administer that intervention in exactly the way that minimizes its devastating potential for harm so that the cure won't be worse than the disease. There are interventions, like immunizations, that are somewhere in between, in that they do a lot of good, are widely indicated, and therefore are amenable to engineering so that you don't need a professional to decide to apply, or to then actually administer, the intervention. If an intervention wasn't harmful, and did not have a very thin margin between its risks and its benefits, it wouldn't be limited to licensed providers, so no, it isn't surprising that bad outcomes and professional medical interventions correlate, and professional medical interventions do not show a huge therapeutic margin.

Now, this consideration might be thought not to apply to geographic variability in medical care. Common sense would seem to say that people should be basically of the same health throughout the country. But if your common sense tells you that, it's dead wrong. FL has a hugely higher death rate than AK. FL also has a higher per capita concentration of medical professionals, so, eureka, people die more in FL because of all those medical professionals imposing their uneccessary interventions!? Well, not really. People get sick, get interventions and die more in FL than AK, because they only feel safe staying in AK if they're in basicly sound health, and they tend to move to FL when their health starts to fail.

You can get rid of most of these statistical differences between AK and FL by adjusting for age. If Drum were quoting medical literature, I would be confident that the statistics in question that show flat outcomes despite big differences in intervention rates had been adjusted for age. But the medical economics literature isn't very intellectually fastidious, so I would have to review his sources to be confident that they had done even this elementary due diligence. But even if we're generous and assume that the figures have been properly adjusted for age, gender, perhaps even variable rates of underlying illness, you can't adjust for things you can't measure. I would maintain that, even matched for age, gender and diagnoses, people who are more threatened by illness in ways that the statistics can't capture, are more likely than someone of same age, gender and even diagnosis, who just plain isn't as sick, to take the precaution to move into or near the big city, or to the Northeast, or to FL, where the full panoply of medical interventions is available. If these medically better-served areas of the country are able to, medically, run fast enough just to keep up with areas that are probably less burdened with illness, then the extra interventions are hardly pointless.

One of the benefits of a well-designed single payer system is that it would have a very powerful effect towards getting medical services better distributed throughout this country. Then maybe people would be under less pressure to move to another state when they get old and sick.

Posted by Glen Tomkins | September 4, 2007 5:34 PM

It's really kind of a perfect isomorphism with the problem of Democratic political consultants getting paid with a percentage of the money they spend to implement ad buys, isn't it? Perverse incentives all around.

Posted by Brodysattva | September 4, 2007 7:00 PM

Speaking as a health care professional, I'm all for this. I ran screaming from private practice, due to the insanity of managed care, a few years ago. I now hold 2 salaried positions. I'm making more money than I was before, I have access to employer benefits, and, best of all, I no longer have to deal with the business end of health care. I can allocate my services to my patients' needs, without thinking about what's billable and what isn't. It's a shame so many health care providers don't consider this more seriously.

Posted by beckya57 | September 4, 2007 8:07 PM

Re: But the theory of an HMO -- if certainly not the practice -- is hardly radical and it is exactly what MY seems to be getting at here and is actually exactly how health care should work

HMOs were originally a good deal for patients too. They were designed to be non-profit, to use community rating (they accepted everyone during open enrollment), and friendly to individual suscribers. Their premiums were low, and their copays low or zero. The only real problem (though not trivial) was the lack of coverage when traveling in areas outside the HMO network, which tended to be geographically small as HMOs were originally quite localized. When they were first introduced they were seen as a way to bring about universal healthcare since they would cover people who did not have health insurance at work. Then employers discovered their low costs and shifted milliosn of employees into them, which strained their networks and ran up their costs and turned them into employer group plans, and then the for-profit insurance companies got into the act, and then Insurance Money bribed state legislatures to drop most of the regulations that had made them consummer-friendly and then- they turned into the monstrosities that everyone hates today.

Posted by Jonf | September 4, 2007 8:59 PM

I couldn't help being reminded of this SNL skit from the Belushi Days:

http://snltranscripts.jt.org/78/78fmercy.phtml

Doctor: All I'm saying is that she has terminal metastasis of the liver and vertebrae and will never be capable of walking, moving or communicating. And because of a total deterioration of the vestibulo-colear nerve, she is, however, capable of experiencing excruciating pain.

Mrs. Gilbert: Oh ...

Mr. Gilbert: [puts a comforting arm around his wife] I don't wanna sound cold about it but, uh, two thousand dollars a day to keep an old woman who's had a full, happy life in excruciating pain doesn't sound like much of a bargain to me. ...

Posted by Njorl | September 5, 2007 12:44 PM

In my experience, much of this overtreatment stems from the increasing specialization of the medical field. For instance, a patient's cardiologist will attempt to get her cholesterol down to 100 or 120 in order to reduce his cardiac risk, without considering that she needs cholesterol in order to make vitamin D, and steroid hormones like estrogen, progesterone, and cortisol. Her risk for mental illness, cancer, osteoporosis and other problems go up. Or they'll try to get an 80-year-old's blood pressure down to 110/70, which makes the old girl dizzy and unsteady, so she falls and breaks her hip, which has a mortality rate of 50% for the elderly. They don't seem to focus on anything other than the risk factors in their own field.

There needs to be some organization of the system wherein a primary care doctor or some other non-specialist is responsible for looking at the big picture, and the patient's preferences for quality of life vs. risk management need to be considered. Supposedly things are organized that way now, with PCP's as "gatekeepers," but the reality is there's little to no communication between health care providers because nobody seems to have the time. They're all seeing 30 patients or more a day so that they can pay their bills, since insurance companies will delay and object to reimbursing.

Posted by 14All | September 5, 2007 12:50 PM

there's a lot about medicine that still seems pretty medieval and guild-like, regardless of the technology now available...there was a lot of squawking about 80 hour work week limitations on residents and how that was going to reduce quality of care and training, when it seems that just the opposite happened. (and I don't know about you, but 80 hours is still way too long for me). Doing away with "piecework" and moving to salaries, plus bonuses (like most other professionals) would seem to have numerous benefits. See also, an article in Slate, about a few medical practices moving to "same day appointments" as being the standard for the practice. My doctors all seem to be moving in the opposite direction, unfortunately. While at one point the OB/Gyn was the only place where it was impossible to schedule an appointment less than 3 months in advance, this is starting to trickle over to "maintenence" visits at my doctor and my kids pediatricians--which wasn't the case a few years ago. Something is wrong, clearly, but I guess like any other slow moving large organism with a lot of non-communicating parts, it's hard to get to the bottom of the problem and even harder to change course.

Posted by JMS | September 5, 2007 2:15 PM