A Better Way to Die

May 10th, 2009 by Al Lewis (alewis)

Dying is expensive and inefficient — even as compared to the rest of our healthcare system, which is quite a feat.   One of the tenets of OOBonomics is to reduce “drags” on the economy, of which healthcare – particularly government-financed healthcare – is high on the list.  Medicare spends one-third of its entire budget for the final two years of life for chronically ill people, years fraught with ultimately futile and usually debilitating and/or painful medical care.


Keeping terminally people alive against their will, the epitome of cost-ineffectiveness, is a “tradition” whose time has passed.  Once again, this is OOBonomics, so no one is saying that people can’t be kept alive, no one is rationing end-of-life care, and no one is denying any access to treatments.  Even without any of those three mandates, though, there is huge potential to improve both costs and the patient experience, in at least six ways.

(1)   Make sure that patient preferences are taken into account

Let’s start with “advance directives.”    An individual uses advance directive to review and specify treatment options ahead of time for contingencies in which he or she would not be capable of making a decision on the spot.  An example would be permanent unconsciousness, as might result from a massive stroke or a prolonged period of oxygen deprivation to the brain following resuscitation from a cardiac arrest. The phrase “permanent unconsciousness” includes both coma, in which a sleeping patient cannot be awakened, and vegetative state, in which a patient, despite having periods of “wakefulness” (eyes open, movement) shows no awareness of his or her surroundings, and is therefore “unconscious” of the environment.   In an advance directive, a person would specify whether or not, in the case of permanent loss of consciousness, he or she wanted to be kept alive on a ventilator or a feeding tube.


Advance directives come in different forms in different states. A living will specifies a patient’s wishes in different circumstances, while a health care proxy (or power of attorney for health care) designates a trusted friend or family member to make medical decisions if a patient cannot do so.  Even with both documents in place, contingencies may arise that no one thought of, or are not thoroughly covered, or the “directive” may be ambiguous.  Other times, physicians may disagree on whether a patient is indeed terminally ill.   Imperfect as they are, advance directives are nonetheless a useful means of assuring that a patient gets the kind of medical care he or she really wants. 

A very simple OOBonomics step would be to create financial incentives and disincentives for physicians to ensure that all of their adult patients have correctly completed these documents. (One solution is to increase physician reimbursement for each patient who died in the hospital or hospice with an advance directive or health care proxy on file, and reduce it for each person who died without one. 

(2)   Create insurance plans that don’t cover long-term life support in the event of brain death

What do Botox and artificial life support for the terminally ill have in common? 


Answer:  They are both administered in a health care setting, but neither is health care.  Both are personal choices.  No one would say that I shouldn’t be able to have Botox if I want it and am willing to pay for it.  Why not say the same thing about artificial life support for the terminally ill?   Why does Medicare, often combined with Medicaid when Medicare runs out for people who qualify for Medicaid, cover artificial life support?  Both are personal choices, not health care.  Botox isn’t covered.  Why should life in permanent unconsciousness be covered?


One interesting observation about permanent unconsciousness coverage comes from a USA Today/Kaiser/ABC poll asking people how they feel about artificial life extension.   According to this poll, 40% of us want to keep people alive, and presumably be kept alive themselves, at all costs. But costs to whom?   Taxpayers, that’s who.    The expense is invisible to the actual patient.  So of course some of us want to be kept alive at all costs:  It’s free.  Under that scenario many people would want Botox too.


Why should others be paying their bills for artificial life extension? Healthcare is subsidized both because society has an interest in having a healthy population and because there is an overwhelming consensus that it is not right to deny people basic healthcare – healthcare which keeps them out of pain and makes their life bearable and productive – because they can’t afford it.   However, artificial life extension doesn’t fit either rationale any more than Botox does.


Let’s consider a Terry Schiavo-type case.  The country divided into two camps over her:  those who think she should have been kept alive and those who wanted to pull the plug.  But was there a third way?   A way which would prevent similar cases in the future from ending up in a controversial situation?  Yes — Exclude life support for the terminally ill/permanently unconscious from the basic insurance benefit and allow people to purchase such coverage as a rider if they want it.  Taking life support out of the cost of insurance reduces the cost of insurance for people who don’t want to be kept alive at all costs, while not denying that right to those who do.  Had this rider been available, Ms. Schiavo’s wishes would have been known simply by noting whether she had purchased it.

Absent the rider, maybe the basic coverage ends at the point at which, for example, two physicians not involved with the patient or family certify the extent of permanent injury to the brain.   The odds of recovering from coma or vegetative state become smaller and smaller the longer it persists, and become a virtual impossibility after several months, especially for non-traumatic brain injury.  For example, a guy who goes into a coma after striking his head in a car accident is more likely to recover than if he goes into a coma after his heart stops for several minutes.  Still, at some point OOBonomics becomes medical ethics and I know enough about the latter to know that it is time to shut up. 

Those 40% of people who say they want to be kept alive themselves when there is essentially no hope of leaving the hospital or regaining consciousness could simply elect the extra rider. (Of course, they could change their coverage at an anniversary date, just like with any other insurance policy.)  Prediction:  Many of those 40% won’t want to be kept alive “at all costs” if it’s their nickel.  Being kept alive may not be such a high a priority for them that they will be willing to actually buy insurance to cover the expense of doing it.

(3)   Make dying less profitable for the hospital

I will let someone else do the talking here.  Joseph Sacco MD, who specializes in palliative medicine at Bronx-Lebanon Hospital Center and who also teaches on the subject, recounts some of the difficulties he has encountered in providing appropriate end-of-life care:

A hospital is a business. Like any businessman, a smart hospital CFO will do what he can to maximize revenues. In health care, that means taking aggressive care of the sickest possible patients.  Such situations constitute the highest-revenue non-transplant patient category in all of Medicare, a category known as “DRG 541.”    A patient who spends five or more days on a ventilator and undergoes a tracheostomy and a major surgical procedure falls under DRG 541.  The hospital’s reimbursement from Medicare for DRG 541 approaches a quarter million dollars.

To maximize revenues for patients on ventilators, the hospital needs to perform a tracheostomy[1]  after the patient has been on a ventilator for at least five days, but before “pulling the plug.”  Pulling the plug before the tracheostomy means reduced reimbursement.  However, to maximize profits the patient shouldn’t be lying around on a ventilator without a tracheostomy either.  The ICU days are adding up, but reimbursement is not increasing.

For patients with terminal illness such as late stage heart and lung disease, cancer or cirrhosis, five days on a ventilator followed by a tracheostomy and a major surgical procedure is unlikely to do anything other than prolong dying. At best, patients are extremely unlikely to leave the hospital in any kind of a functional state, and will become critically ill again in a period of weeks if not days.  Yet they are also the most profitable patients.   The hospital gets paid the most not for restoring people to health but for keeping nearly dead people alive.  I know what you’re thinking, “Al, you can’t be just about to say that hospitals and doctors would try to fill beds with the highest-paying patients just to get paid to keep them alive.  Financial considerations would never trump clinical integrity.   That would lead to the ridiculous conclusion that the more hospital beds there are, the more people will be placed in them.  Everyone knows that people are only admitted to the hospital when they really need to be.”

If you’re thinking that’s what I’m just about to say, you’d be wrong.   I am not “just about to say” it.  An actual authoritative source is just about to say it, and all I am going to do is cite that source.  According to The Dartmouth Atlas, which compiles and compares medical statistics by region, the cost of your final year would be almost twice as high in Miami, Florida, than in Portland, Oregon,   In Miami, in the last six months of life alone, you’d have 46 doctor appointments, mostly with specialists.  In Portland, you’d have only 18, and half of those would be with your primary care doctor.

The same is true for days in the intensive care unit:  Miami residents have six, Portland residents only one.  The primary reason for this disparity is, according to this very same The Dartmouth Atlas, that Miami has more ICU beds and specialists than most other places.  Because humans have the same physiology in Portland as in Miami and because life expectancies are similar in both cities, it is unlikely that the difference in utilization of ICU beds and specialists is due to anything other than what the Dartmouth Atlas concludes, which is that more ICU beds and more specialists generate more ICU days and more specialist visits.

So it turns out that hospitals and physicians, like real estate brokers and other people such as you and me, tend to be driven by economics, in this case the economics of DRG 541.   Change the economics to align incentives, and you’ll change their behaviors.  Futile care is like anything else:  Pay for less, and you’ll get less.  Although, once again, let us be clear:  OOBonomics is not advocating the rationing of any form of medical care, end-of-life or otherwise.  OOBonomics doesn’t do mandates.

(4)   Encourage hospitals to focus on helping people die

One habit certain to change once incentives change is the general lack of hospital interest in helping people die.  Most hospitals don’t have specialists on staff to do that.  Doctors like Joe Sacco who do that are called “palliative medicine specialists” or “palliative care specialists.” They help people die just the way obstetricians help people get born.  Births outnumber deaths in the U.S. only by about 1.7 to 1, but obstetricians outnumber palliative medicine specialists by about 15 to 1, despite the much greater medical complexity and counseling needs of most dying patients as compared to most pregnant women. 

Here is a primer on the role of palliative medicine specialists.  Many people who are terminally ill when admitted to the hospital have not been counseled by their own doctor as to what to expect, and many lack advance directives too.  Even those who do have advance directives and have been counseled by their doctors may be frightened and conflicted about what they want as the end approaches. In the absence of a palliative care team, these patients often find their way to the ICU, because specialists whose training and whose reimbursement are both geared to doing more interventions are not as comfortable or as well-paid for doing fewer, or even for discussing realistic alternatives to critical care.  Not to mention that those ICUs need to be filled. 

The palliative care physician’s job is just the opposite – to counsel people on all their options, from intensive care to comfort-only care.  The physician has no financial stake in the patient’s choice.  Pretty consistently, two-thirds elect comfort-only care with perhaps only minor interventions like antibiotics.  The USA Today poll proves not to be far off in practice:  About a third of people who are counseled regarding all alternatives choose the ICU, ventilator, and resuscitation in the event of cardiac arrest — the whole ball of taxpayer-financed wax.  

Either way, the palliative care team (nurses and social workers as well as physicians) helps people avoid futile care if they don’t want it and allows them to choose it if they do want it.  The team also addresses the fact that most terminally ill people don’t know what to expect in the hospital, or helps them when they suddenly change their mind.

Not surprisingly given the amount of money spent at the end of life in medical care and given that theirs is the only specialty whose goal is to provide less care in many cases, the cost-effectiveness of palliative care could be the highest of any specialty.  Once again, I will let the palliative care specialist tell the story:

My palliative care team has had more than 600 instances in which patients signed Do Not Resuscitate (DNR) orders before suffering in-hospital cardiac arrest. All were patients with chronic and progressive illnesses who, had they been resuscitated, would have spent a few days or weeks unconscious on life support before their hearts stopped for good.  That’s 600 pointless cardiac arrest resuscitations avoided, 600 families spared watching their loved ones rot away in the ICU, not to mention maybe 3000 or 4000 needless additional days in the ICU. We showed that we saved the hospital $3 million a year. 

What creates the OOBonomics opportunity is — assuming Medicare is paying for ICU stays and the cardiac arrest resuscitations — that $3 million in cost savings often if not usually translates into lost revenues of much greater than $3 million.  

In general and for most types of insurance, aggressive care, no matter how unwarranted, is more profitable for the hospital than conservative or palliative care in almost all circumstances involving terminally ill people.   It costs much more too, but the patient isn’t paying for it directly, so it is not surprising that a third of patients say: “Bring it on.”

Perhaps reimbursing DRG 541 at a lower rate should be offset by rewarding hospitals for moving people into hospice settings and humanely managing the dying process, even if it means providing fewer units of care, however measured. While paying hospitals to provide less or even no medical care (in the traditional sense of the phrase “medical care”) may seem radical, there is precedent:  Some utilities are now rewarded for helping customers to conserve energy and reduce power bills rather than for building new generators.   You heard right.  Utilities can get paid more for providing less power when less power is cost-effective for buyers.  Why can’t hospitals do the same, when less futile care is cost-effective for buyers (meaning Medicare) and preferred by patients?

Aligning financial incentives with the actual wishes of patients and their families should make more facilities more conscientious about responding to those desires.  Of course doctors and hospital administrators will all be as horrified to hear that observation as real estate brokers were when I posted my essay on ThinkOOB about their commission-based incentives.  Medical and hospital staff would say that they already respond to patient wishes and would never put their own revenues ahead of ethics.  In that case they shouldn’t mind this proposal at all.  I’m sure every single one of them is making the best efforts already to find those advance directives, and never pushes their patients to undergo heroic one-in-a-thousand-chance-of-cure treatments instead.  That’s why the following joke is not the slightest bit funny:

Q: Why do coffins have nails?
A: To keep oncologists out.

We noted in the real estate chapter that if realtors genuinely thought they were offering clients a valuable service, they wouldn’t mind having their fees become market-driven.  The same is true of this advance directives proposal.   In both cases, the fact that there is opposition makes my point better than I can.  Having said that, the opposition to the advance directives proposal, as measured by total comments, was considerably more muted than to the real estate proposal.

(5)   Bring death to life

Hospitals are required to report many variables related to the quality of care, such as infection rates, readmission rates, death rates in certain types of surgery, and so on.  None of those variables relate to how well they handle dying patients.  Yet there are certain variables that correlate very closely with quality-of-death that should be brought out into the open.  How many people died after more than two weeks in the ICU?  How many died after one, two, or more cardiac resuscitations?  What is the hospice referral rate?   How many people who died had advance directives in place?  How many people had multiple resuscitations from cardiac arrests?  What were the satisfaction scores for the families of patients who died in the ICU?  How many people on ventilators had tracheostomies and subsequently died in the hospital?

(6)   Change the “default” provision in the event no advance directive is signed

The simplest proposal and perhaps the one with the greatest savings of all would be to change the “default” provision for end-of-life care.  Remember, only 40% of people want to be kept alive at all costs (and that is before the alternatives are explained, which in at least Dr. Sacco’s hospital reduces the proportion to 33%), so why do we assume that absent a directive, everyone does?  Protocols for most medical treatments are so widely followed and well-established that they are used as a defense for malpractice claims today.  The law reads something like this:  If a plaintiff (patient) alleges that care was substandard and a physician can show that the care was provided according to a published standard, the burden falls on the plaintiff to prove substandard care.  If the care was not provided according to the published standard, the burden falls on the physician to show a medically sound reason for the deviation.

Likewise, why can’t there be a default protocol in the event of terminal illness?  Patients could deviate from the protocol with their own advance directive but otherwise it would be the job of the physicians to do exactly what they do in all other areas of medicine, which is to apply the standard of care unless in their professional judgment there is good reason to deviate from that standard.  Given that only a minority of patients wants to be kept alive at all costs even when they aren’t paying for it, and given the huge expense to the system in end-of-life care, and given the atrocious quality of life for terminally ill people kept alive by artificial means, it is likely that the task force assigned to develop that protocol would agree on a default standard of care that would look much more like a typical completed advance directive rather than a typical experience today.



Share This Idea:
  • Digg
  • StumbleUpon
  • del.icio.us
  • Facebook
  • Mixx
  • Google


Tags: , ,

16 Responses to “A Better Way to Die”

  1. thinkoobfan Says:

    best I’ve ever read on the subject. Be prepared for some hate mail

  2. harvardeconomist Says:

    your best posting ever. I could quibble that just not covering unconsciousness wouldn’t much money. You should say that basic insurance should only cover comfort care for all terminally itt

  3. reader Says:

    pretty radical but a big improvement on your earlier one. this is what they should do

  4. actuaryhater Says:

    Milliman just put out a report on how to save money in healthcare and it was full of healthcare cliches. This is much more interesting. You should send it to the White House.

  5. radiostar Says:

    I just heard you talking about this stuff on the radio in Atlanta. I am surprised you didn’t get hacked or deluged with hate mail. Tell those trailer scum I am tired of paying for their health care and this is a good place to start. you’re right-this is NOT health care. it’s religion and the state shouldn’t pay for it

  6. canurse Says:

    Your interview on Montel aired here a little while agoi, and I want you to know that not all Cailfornia nurses sound like Jeri. I am in the nurse association here and I am totally embarrassed by what she said. you were right to take issue and I’m glad you and Montel asked her to leave. my apologies.

  7. califnursenancy Says:

    I will not be voting for her next time she runs for president of our association. she made all Calif. nurses look like selfish whiners. you and Montel were rather respectful under the circumstances

  8. toby Says:

    My grandmother died recently. She had a living will but no one could find it and was resuscitated twice. If this had been the law that would not have happened

  9. doodler Says:

    I heard you on the radio too. I didn’t get what disease management is and what it has to be with the single-payor system and you never said whether you thought Montel’s idea for a tax on unhealthy food was a goodf idea. Can you cooment?

  10. alewis Says:

    Montel’s idea for a tax on unhealthy food is a great one in theory but the devil is in the details. Health care could appear in a THesaurus next to “law of unintended consequences.” Among other things, the tax would be very regressive. You’d have to offset it with subsidies for other kinds of food. Also there is much debate over what constitutes healthy food and the definiton changes over time. Dark chocolate and eggs used to be no-nos and are now considered good for you.

  11. eyechart Says:

    There’s no hate mail because religious people can still prolong life. this proposal doesnt’ take that right away. It is hard to argue with something that doesn’t take away something else.

  12. Thaddeus M. Pope Says:

    Great post. I focus a good deal on the relatively few cases where patients really do want aggressive end-of-life care contrary to provider recommendations.

    It is good to be reminded that patients and surrogates are not the cause of suffering and high costs nearly as much as providers are (given current incentives).

    To some extent, we do not even need NEW tools and incentives. For example, advance directives are decades old. But they are not completed, not located, not clearly written, not understood, and not followed even when understood.

  13. Thaddeus M. Pope Says:

    I would suggest one modification. Instead of “Pay for less, and you’ll get less,” write Pay for less, and you’ll get MORE.”

  14. thinkoobfan Says:

    can you write something on the CAFE standards? It’s time for a new posting anywqy

  15. cvb Says:


  16. Hospice Los Angeles Says:

    You made some good points there. I think most people will agree, I’m going to look into this some more - thanks! Thanks for the info. A good find. Your article was very interesting. Thank You.

Leave a Reply