Wednesday, April 13, 2011

Health Care Debate

The US is undergoing a national debate about health care.  This debate tends to be dominated by politics rather than by analysis of evidence. Health care is indeed a complicated issue.  I would like to make a few observations and recommendations.

1) The US health care system is not the best.  We spend nearly double what other developed countries spend on health care and our health care costs continue to explode. Compared with other developed countries we obtain inferior results in terms of objective outcomes such as longevity, infant mortality, and management of chronic diseases.  Some individual health care facilities and practitioners in the US are arguably the best in the world, but as a health care system our results are far from satisfactory. The first step in improving a situation is recognizing that a problem exists.

2) We could as a society decide that only those individuals who can pay for health care, either because they have purchased insurance or because they can pay out-of-pocket, should receive health care.  If we made this choice, individuals who could not pay for health care would simply go without it and either recover on their own, get worse, or die.  Fortunately, it appears that there is a consensus that as a society we do believe that people who are sick should be cared for even if they do not have means to pay for their care.  This consensus is based primarily on empathy.  Of course, there are pragmatic reasons as well to keep people healthy and care for the sick.  If contagious diseases are not prevented nor treated, all of society is at risk.  Also, people who are sick are less likely to be able to work and thus contribute to the economy and care for their dependents. A nation benefits from a healthy population.

3) At the same time, we should recognize that there is no limit to the amount of money that we can spend on health care.  New diagnostics and therapies are continually being developed.  For many of these one may be able to demonstrate some clinical benefit.  If a third party payer, such as Medicare, pays for all such new diagnostics and therapies the cost will continue to spiral upward.  In fact, by statute Medicare currently cannot deny coverage for a procedure on the basis of cost as long as the procedure has been demonstrated to provide some clinical benefit.

4) Health care costs are spiraling up for private third party payers as well as for government insurance programs.  Physicians and patients have an incentive to consume more health care as long as a third party is paying for it.  Insurers, whose profit represents a percentage of their revenues, also have little incentive to control health care costs as long as employers are willing to pay the premiums.

5) If we believe that everyone should have access to health care, but that we cannot afford to provide unlimited health care, then we need to devise a system that achieves a basic level of universal health care with a cap on the expense.  If funding is limited, health care - at least that health care provided by taxpayer funds - should be spent in the most efficacious way.  I would propose that we eliminate Medicaid and  Medicare as separate programs and let the government fix a level of expenditure (perhaps expressed as a percentage of GDP) that it will dedicate to providing basic health care for all.  There are several ways that this could be implemented, for example:

a) One option  is that we adopt a national health service like the one in England where the government owns the health care facilities and employs the health care providers.  Coverage for health care services would be determined on the basis of relative cost-effectiveness in order to keep the expenditures within budget.  For example, different treatments for the same condition would be compared on the basis of relative effectiveness and cost, and only the most cost-effective ones would be covered.  Even the most cost-effective treatment for a given condition might not be covered if compared to treatments for other conditions it does not meet a cost-effectiveness threshold.  For example, a million dollar treatment that is expected to prolong the life of a cancer victim by two weeks would undoubtedly not be covered.

b) A second option is that the government becomes the single payer for all health care services, but the health care services are provided privately (Canada's system).  This option is also termed  the "Medicare for All" system.  Under this option as well, coverage for health care services would need to be determined on the basis of cost-effectiveness in order to control costs.

c) A third option is that the government provides everyone a voucher for basic health care and individuals use it to purchase health care insurance (this is something like the Paul Ryan plan for Medicare).  All plans would have to meet certain government standards (e.g. coverage cannot be denied for pre-existing conditions, out-of-pocket caps on co-pays, specified diagnostic procedures and therapies demonstrated to be clearly cost-effective must be covered, etc.)  Individuals who do not sign up for a plan would be assigned to one.  The voucher payments to the insurers would be adjusted based on the expected health care costs of the individuals they cover (e.g. actuarial adjustments based on age, current illnesses, etc).  I would propose (unlike Paul Ryan's plan) that the basic health care plan be fully paid for by the voucher. This would force insurers to come up with basic health care plans that fit within a fixed budget.

6) In principle, I would favor the voucher option because it should allow for greater innovation in health care as a result of competition among insurers and health care providers.   For example, it may be that some health care facilities take over the role of insurer and directly provide the benefits thus eliminating the insurance middle man.  Some plans may emphasize preventive care and thereby reduce the costs of chronic diseases, etc.  The voucher plan also makes it possible for the government to fix its health care expenditures at a predetermined level.  This plan, like the other options, separates health care from employment so that individuals do not have to change their health care plan when changing jobs and also frees employers from having to provide health care insurance.

However, I would not be in favor of implementing any plan immediately on a nationwide basis.  We should prospectively test candidate plans first in local areas (e.g. cities, counties or states) to see how it performs before implementing it nationwide.  Health care is far too complex that anyone can anticipate all the issues that will arise with any specific plan.  Prospective testing is needed to obtain evidence regarding performance.

7) The above plans would cover basic health care.  Insurers could provide supplemental plans that provide additional coverages or amenities (e.g. private hospital rooms, etc). Individuals would pay privately for the supplemental coverage.   We allow individuals to buy yachts, provide private education for their children and we should allow individuals to pay for additional health care services.

8) In summary, I recommend that we should simplify our health care system by providing a basic level of health care for all, fix the taxpayer expenditure that will be dedicated to this goal, spend the available taxpayer funds in the most cost-effective manner, merge the multiple government health care plans into one plan for all, and separate health care from employment.  Candidate plans to meet these objectives should be first prospectively tested in local areas prior to being implemented on a national basis.


  1. It may be a bit unusual to add a comment to one’s own post, but I wanted to address one possible objection to the voucher proposal that I had suggested without putting up a new post.

    The objection is that the voucher proposal would necessarily result in an increased overall federal budget. Currently, health care consumes 17.5% of GDP. The government (federal, state and local) already funds approximately one-half of health care costs in the US. If for argument’s sake 11.5% of GDP were spent on the new vouchers for basic health care, the government would have to contribute approximately 2.75% of GDP (approximately 400 billion dollars annually) more than it is already spending (assuming all government health care programs were replaced by the universal voucher). There would be additional non-government health care expenditures coming from co-pays, supplementary insurance programs and services which are provided on a self-pay basis (like cosmetic surgery). The expectation would be that the overall percent of GDP spent on health care would come down because of competition among insurers to provide basic health care within the fixed budget of the voucher system.

    Whatever the specific amount is, the government would be contributing significantly more than it currently is to health care and would require tax revenue to support that expenditure. However, this is primarily a book-keeping issue because the money for basic health care would be returned to citizens as a voucher that they would use to buy private health insurance that the great majority of citizens currently purchase anyway.

    To prevent overall government expenditures on health care from nominally increasing one could attempt to needs-test the vouchers so that the poor are subsidized and everyone else pays the full cost of the basic health care insurance directly to the insurer. In this scenario, everyone would be required to purchase, at a minimum, the same basic health care insurance that the poor can purchase with the voucher. One problem with this “individual mandate” approach is that its Constitutionality is in question – the Federal government may not have the authority under the Commerce clause of the Constitution to impose a mandate on individuals to make a specific purchase.

    Even, however, if one could get around the Constitutionality issue by replacing the mandate with a tax penalty for not purchasing the insurance (or tax credit for purchasing the insurance), another problem remains. How would the basic health care insurance be priced? The government vouchers would have an average dollar value. However the actual payments to the insurance companies would be adjusted on an actuarial basis to compensate for the variation in cost of care based on the demographics of the population that each company covers. These demographics include age, gender, and prevalence of various medical conditions. This is very important because in the absence of these actuarial adjustments insurance companies would be incentivized to “cherry-pick” those beneficiaries whose costs of care would be expected to be lower than average. However, if individuals paid directly for basic health care insurance, and we did not allow pre-existing conditions to affect the price of the insurance, then the insurers would be indeed be incentivized to cherry-pick beneficiaries. Conversely, if we allowed pre-existing conditions (including age and gender) to affect the price of insurance, coverage costs would escalate for the old and the sick.

    I don’t think that it would be impossible to devise a system where most people self-pay for basic required health care coverage and others were subsidized, but I think such a system would be more complicated, and more readily subject to abuse, than a universal voucher system. It would still hold the promise to be far superior to our current system. Politically, it may be much more feasible because it could actually reduce nominal government expenditures on health care rather than increasing them.

  2. Dr. Cohen

    I am interested in the voucher system presented above. I am curious as to why the voucher holds monetary value if the burden of purchasing a plan is placed on the uninsured.

    I agree with the unconstitutional nature of forcing a U.S. citizen to purchase anything, let alone an insurance plan. However, perhaps we can provide incentive for the uninsured to purchase plans. Everyone, at some time or another requires healthcare and a medication of some sort is almost always dispensed. If a medication costs $10 on insurance and $200 without insurance, then why wouldn't a company require health insurance for all drug related purchases? They have no problem enforcing rules that allow for cherry-picking, as is mentioned. This behavior violates human rights in my mind. I believe insurance companies must share the burden of the uninsured just as physicians and hospitals do.

    For example, a low budget plan aimed at low-income families could be available via insurance plans. Subsidizing healthcare is the other option; however, how does this prevent unemployment compensation-like behavior from happening? Knowing one can obtain subsidized healthcare without having a job perpetuates the cycle of government funding. It creates a one-stop-shop for charity care. A healthy population is one that is more willing to work in my mind.

    We do not usually turn patients away, although I have heard of patients being rejected for complex surgical cases. This can be thought of as clinical judgement. But, allowing for people to pay out of pocket for premium services gives the physician unfair incentive to perform complex surgical procedures without clinical incentive. Physicians are not void of human behavior.

    Insurance companies that dictate our behavior as patients and physicians is a major dilemma of perspective. Profit driven medicine must match life-driven medicine. In effect, I believe the salary of the individual should dictate the price that should be paid for coverage. It is already being done by taxation policies. And, this would force the insurance companies to stop dictating our behavior to maximize profits.

    Ethical dilemmas present with the current state of profit driven medicine. For example, Dr. Peter Carmel, President of the AMA has presented graphs whereby the level of physician compensation has been declining while the number of administrators has gone up. It is unclear to me where the money is going given that the monetary pipeline of healthcare is not open-source. But, the funding may also be going towards paying for the uninsured. As a hospital accumulates debt, the state is required to pick up the tab if it is a state institution. So, an uninsured patient comes in, gets treatment, and the state pays, correct? Meanwhile, there are tiers of administrators piling up to process the paperwork involved. I do not have numbers on this and only know it through intuition. Perhaps you may be able to shed some light on how the Mass state and private hospitals deal with this problem.

    I have read your post with great interest and look forward to hearing more.

    Dr. Jekyll

  3. Vaibhav Devidas Patil, M.D.April 18, 2011 at 10:48 PM

    Professor Cohen,

    I am curious to see why the vouchers mentioned have monetary value. I do not understand, if the burden of purchasing the insurance plan is on the patron, why there is a monetary value associated with the voucher.

    I agree that forcing a U.S. citizen to purchase insurance could be perceived as unconstitutional. I also believe that the plans might be based on how much an individual makes per year to sidestep this dilemma. If a patient requires care and pays $10 for medication without insurance but $200 with insurance, then perhaps a low budget insurance plan would suffice. This would ensure that all medication dispensed has been covered by insurance and, in effect, that a larger population is insured. This might even be accomplished by convincing insurance companies to invest in smaller thread like companies in underserved areas. As they operate on a small scale within these underserved areas, the larger firms would be able to deal with larger populations simultaneously while coaching them into the larger scene. This might take the place of the subsidized plans.

    This might require a few modifications to the types of coverage that insurance plans offer. Namely, they may have to change policies such that there is no discrepancy in care depending on the type of coverage. I believe that cherry picking, although it makes sense from a profit based perspective, violates human rights. The patient population should be able to choose what type of care they receive as long as the caregiver is comfortable with delivering the care.

    I realize this is a common sense perspective and likely argues against your evidence-based perspective. However, I would argue that once an evidence-based theory becomes a law, it is perceived as common sense. And, much of medicine, including many aspects of psychiatry has moved leaps and bounds using a common sense approach to illness. Insuring the U.S. population is an emotional decision on many levels. Certain groups will undoubtedly feel neglected and underrepresented. It makes sense to me that we should rely on our field of medicine that has walked these lines in the past and made strides, namely psychoanalysis in psychiatry. As much of this field progresses into the era of evidence based medicine with psychopharmacology, there seems to be a positive correlation between the human component and the scientific component. I do not agree that evidence trumps common sense. I believe they work hand in hand.

  4. Doesn't the problem generally come in at the time when "basic health care" has to be defined (what must be covered by the voucher)? Especially at the end of life? You bring up the $1 million for 2 week life extension earlier in your post - would that be covered under the basic healthcare plan? What about $2 million for 6 months? What about a 6 year old with a chronic condition that costs a hundred thousand dollars a year to treat, but with continuous treatment, he will live a full and normal life?

    Make "basic healthcare" too inclusive and no insurance company will be able to provide a plan at the voucher cost - and you can't make a insurance company participate in the system - which I think would quickly become an obvious issue with Paul Ryan's plan, especially at his voucher worth growth rate.

    Or the voucher cost would quickly become very expensive for the government.

    But if you don't cover everything, there is always a very interested patient group that will make very loud and sympathetic noises to Congress about it. And where would the pushback come from?

    Also, isn't part of the goal of the ACA to use Medicare's enormous buying power to force changes in how we pay for care? If Medicare is disbanded and split into many insurance plans, what would replace this method of cost containment? Or do you not believe that this will be effective anyway?

  5. This comment is in response to Dr. Jekyll (and fortunately not Mr. Hyde), Dr. Patil and Chris. Thank you all for providing your comments and participating in the discussion.

    Drs. Jekyll and Patil raise a number of questions about the voucher system I proposed. In short, the proposed voucher system for basic health care is in response to the view that we should provide some level of health care to everyone whether or not they had the means and foresight to purchase insurance and the hypothesis that it is better to manage diseases prospectively rather than just deal with them when a patient presents to the emergency room for urgent care. The proposed voucher system enables the government to predetermine its health care expense by specifying the average dollar value of the voucher. (To help minimize the incentive of insurers to “cherry-pick” healthy patients, the actual payments to insurers would be adjusted on an actuarial basis based on the demographics of the population they insure. So, for example, an insurer that covered mostly healthy young patients would receive less per voucher than an insurer that covered mostly old sick patients.) The voucher system would enable each insurance company to innovate in terms of how they deliver care and manage diseases subject to the constraint of some government requirements (e.g. caps on out-of-pocket patient expenses, no exclusion for pre-existing conditions, coverage of procedures widely agreed to be indicated - e.g. appendectomy for patients with acute appendicitis). Each insurer would be incentivized to reduce costs by bargaining with health care providers and drug and device manufacturers and specifying which procedures are covered (subject to government requirements) in order to provide the most attractive health care package within the funding level provided by the voucher. Plans would compete to attract patients, and third party reviewers (e.g. Consumer Reports) could provide comparative evaluations to aid in this process. Patients who desired coverage beyond the basic health care plan could purchase supplementary insurance.

    Chris raises an additional issue. What if the government imposed requirements for basic health care are so comprehensive that they cannot be provided for within the voucher payment? The intent of the plan is that the government would impose requirements only for therapies which are clearly indicated (e.g. the six year old with a chronic condition that costs one hundred thousand dollars per year to treat but which treatment enables that individual to live a full and normal life) but leave significant latitude to local insurers for how to manage less clear cut situations. No system is foolproof, however, and if under political pressure the government requires more than can be accomplished within the funds provided by the vouchers, then insurers would begin to withdraw from the market and hopefully provide sufficient push-back to rectify the situation.

    No system is perfect, and any of the systems I proposed will have problems associated with them. That is why any candidate system needs to be prospectively tested and evaluated in local areas before being implemented on a nationwide basis.

    A final comment to Dr. Patil: common sense may at times be a perfectly reasonable means of formulating a hypothesis, but prospectively obtained evidence is required to test that hypothesis.

  6. "if under political pressure the government requires more than can be accomplished within the funds provided by the vouchers, then insurers would begin to withdraw from the market and hopefully provide sufficient push-back to rectify the situation."

    I think that we see with the tax cut extension last year and the yearly doc fix (although neither is a perfect analogy) that, while there would be push-up from the insurers (and people concerned about the lack of insurance options), the pushback wouldn't be for reducing coverage. It would be for raising voucher values. Politically, that would always be the simplest solution.

  7. Chris, you are correct to the extent that any system one can devise will be subject to political influence. In the end the decision on what fraction of GDP the government should spend on basic health care is a political decision.

    However, in the proposed voucher system the pressure to increase the voucher payments would likely occur only if the government body responsible for setting the basic health care requirements established so many mandatory requirements that they could not be covered within the funding provided by the voucher. The insurers would have incentive to negotiate down payments to providers in order to make a profit within the funding provided. This provides a mechanism for controlling costs.

  8. For other countries, health insurance plans can only be obtained by those who can afford them, as you said in your second point. The fact that this isn't true in our country sets us apart.

  9. We must step up and engage in this most crucial debate. Nationalizing our health care system is a point of no return for government interference in the lives of its citizens. If we go down this path, there will be no turning back. Ronald Reagan once wrote, “Government programs, once launched, never disappear. Actually, a government bureau is the nearest thing to eternal life we’ll ever see on this earth.” Let’s stop and think and make our voices heard before it’s too late.