Analysis Essay

Why the US healthcare system is failing, and what might rescue it

BMJ 2012; 344 doi: (Published 09 May 2012) Cite this as: BMJ 2012;344:e3052
  1. Arnold Relman, professor emeritus of medicine and social medicine, Harvard Medical School, Boston, Massachusetts, USA
  1. arnoldrelman{at}

The US healthcare system is by far the most expensive in the world, but it now leaves about 50 million of its citizens totally without coverage and fails to provide adequate protection for millions more. And the quality of care is on average inferior to that of countries that spend much less. Can the rise of multispecialty groups make US medical care much more affordable and efficient?

The US healthcare system seems headed for bankruptcy because of its ever increasing and unsustainable costs. These costs will be effectively controlled only by legislative reforms in the insurance and payment for medical care, but the prospects of such legislation will depend on a more favourable political climate and stronger public support. However, legislation will accomplish little unless the organisation of medical services also changes. The recent movement of US physicians into large multispecialty groups suggests that this reorganisation of medical care may already be under way. If this trend continues, it could not only facilitate the enactment of legislation, but also help to make our medical care much more affordable and efficient.

At present, most US physicians are in solo practice or belong to small, single specialty partnerships, but new social and economic forces are beginning to make employment in large multispecialty groups a more attractive option. About a quarter of all US practitioners are now employed in such groups, which are being formed by independent physician organisations and by hospitals. If their number continues to increase and if they eventually represent the great majority of practising physicians, a wave of legislative reforms could be initiated that transforms our currently dysfunctional healthcare system. But before I say more about this move towards group practice and how it might improve prospects for reform, readers need to understand how and why our health system has been failing. I should also describe the extensive health legislation enacted by the Democrats in 2010, and explain how it falls short of providing the kind of major reforms that would be required to rescue our healthcare system.

The US system is a confusing hodgepodge that makes no sense to informed observers. Although it is by far the most expensive in the world, it now leaves about 50 million of its citizens totally without coverage and fails to provide adequate protection for millions more. Most people also have no support for the cost of long term or rehabilitative care. And, despite huge expenditures, the quality of care is highly variable—but on average inferior to many advanced countries that spend much less. The best physicians and hospitals and the best care available in the US are among the finest to be found anywhere, but access to that care is grossly uneven. Many receive substandard care and far too many receive virtually no care except when illness is far advanced or there is an emergency.1 2

Why the huge cost and the inefficiency?

What explains this paradox? Why do we spend so much on medical care without anything like a commensurate return, and how does our system differ from those in other countries that spend barely half as much but do so much better? In my view, by far the most important explanation—the primary cause of our unique problems—is that the US, alone among all advanced Western countries, has allowed its healthcare system to become a market and its physicians to behave as if they were in business.3 In the US medical care has become a huge, competitive industry with many private investors, but with relatively little government regulation. Involving more than $2.7 trillion (£1.7 trillion; €2 trillion), the US healthcare industry now constitutes nearly 18% of our entire economy and it continues to expand. Its growth has slowed during the past two years, largely reflecting the effect of the recession in reducing employment based insurance and the ability of most people to afford care that is not mostly paid by insurance. Expenditures will probably resume their rise when the economy improves and new federal commitments to pay for care are implemented in the years ahead.

The US is a clear example of what happens when medical care becomes a commodity in trade rather than a social service. Because the market for medical care differs so much from other markets, classical market forces do not exert their usual control over buyers and sellers and hence do not regulate supply and demand.4 In the US medical market, those who largely control the supply of services (physicians) also largely determine the demand. Of course, the patients’ complaints, illnesses, and injuries also play an important role, but it is the physician who interprets them and knows what services are needed for diagnosis, treatment, and prevention. It is also the physician on whom the consumer (that is, the patient) relies for advice. The fee for service payment system through which most physicians are paid gives them a strong economic incentive to be overly generous in their recommendations and provision of services, especially since they are well aware of their reimbursement for the services they provide, but are usually unfamiliar with the charges made to insurers by laboratories, hospitals, and other facilities for the services they recommend.

These facilities, whether paid for each individual item of service or for an episode of care (for example, a day in the hospital or a visit to a clinic), also have strong economic incentives to maximise the volume of their profitable services and increase the referral of paying patients. So, in addition to direct marketing to consumers, they encourage doctors to use their facilities and to refer patients by offering many kinds of favours and financial arrangements. Medical products (prescription drugs, devices, and equipment), which account for more than 15% of costs, are also promoted by manufacturers like commodities in trade rather than being used only according to medical need. Marcia Angell’s much discussed 2011 article documents this behaviour by manufacturers of psychoactive drugs.5 To increase their sales, manufacturers depend on direct marketing to patients, and give financial and other inducements to the physicians who prescribe them. While not unique to the US, these practices are more pervasive in the US than elsewhere. The conflicts of interest that stem from attempts by manufacturers to influence the behaviour of physicians add to the unnecessary costs of the system.

Not only is there little or no government regulation of the volume of services and products in the US healthcare market, but there is also very little regulation of prices—far less than in most other advanced countries. The commercial competition encouraged by this lax regulation affects the behaviour of all players in the market. No more than half of the US health economy involves investor owned organisations and institutions, but most of the others (so called not-for-profits) also see themselves as businesses competing for market share, so they act very much like their for-profit, investor owned competitors. Virtually all organisations and many physicians seek to maximise their income. The net result is the virtually unrestrained growth of health costs, driven not simply by medical need but by economic incentives.

In addition to the cost of a medical care system driven by these economic incentives, there is the huge cost exacted by the dependence of the US system on private for-profit insurance plans. Numbering in the hundreds, but increasingly being consolidated within a relatively few giant corporations, these private plans insure or provide billing and collecting services for more than half of the total population. This includes most of those covered through their place of employment and those who are self insured. In addition, about a quarter of those over 65 have opted to have Medicare pay for their care through private plans, apparently in response to vigorous marketing by private insurers who offer a few services not available in Medicare. Altogether, private insurance plans comprise a huge and growing industry, with a gross income of more than $800bn. Their profits and business overheads vary considerably but average between 15% and 25% of their premiums. The US government estimates private insurance plans added over $150bn to the cost of healthcare in 2011.6 (The overhead expenses of Medicare are less than 5% of total expenditures.) Despite their claims to the contrary, these plans add little or nothing to the value of the insurance they sell or service that is even close to their added cost.7 No other country is as dependent on relatively unregulated private for-profit insurance plans as is the US. Other advanced countries, such as France and Switzerland, include private insurance plans as a central part of their health system, but these plans are not-for-profit and are much more tightly regulated by government than in the US.

The costs of both halves of the US healthcare system—the publicly insured and the privately insured—have been rising at an unsustainable rate, but over the years the privately insured half has risen slightly more rapidly. Despite its much lower overheads and its not-for-profit organisation, the rising cost of public insurance has continued to follow the inflationary trajectory of the private sector. This rise is explained in part by the ageing of the population and the increasing number of citizens reaching the Medicare eligible age of 65. But the chief reasons are that the payment methods and the economic incentives of doctors and hospitals are largely the same throughout the system.

Furthermore, since both public and private insurance plans function largely through billing by fee for service, they are both open to fraudulent billing and inflated charges that enable unscrupulous providers to game the system and extract larger than justified payments. Estimates often made in the media by experts place the cost of fraud and abuse at about 5% to 10% of the total cost of personal care. Another factor driving costs in both publicly and privately insured sectors of the US system is the fear of malpractice litigation and the so called defensive behaviour of physicians who order many tests and procedures to protect themselves against possible lawsuits if anything should go wrong. There are no reliable estimates of how much unnecessary medical service results from defensive practice, but most experts agree it may be a significant but not major contribution to health costs.

The Affordable Care Act of 2010

Although the crisis in the US system is primarily caused by its costs, most of the healthcare legislation enacted in the past few years has been concerned with extending and protecting insurance benefits. A landmark bill, the Affordable Care Act (ACA) was passed by the Democratic controlled Congress in March of 2010. An excellent, readable summary of this huge and complicated legislation can be found in a recent book by McDonough.8 Democrats have acclaimed the ACA as a major step forward in solving the US system’s problems, but many liberals, like me, have reservations.9 The act does promise to cover more of the cost of drugs, to offer subsidised private coverage to about 16 million uninsured citizens, to help states extend Medicaid coverage to another 16 million and prevent private insurers from denying coverage to those with pre-existing illnesses or dropping coverage of those who develop expensive illnesses while being insured. Most of these benefits are not scheduled to start until 2014—although a few have already begun. The law does contain major advances but, despite its name, it has no provisions that will reliably control rising costs. To persuade the politically powerful private insurance industry to support (or, at least, not oppose) the passage of this legislation, President Obama had to agree to include in the law a requirement that all uninsured individuals buy private insurance or be fined—the so called insurance mandate. He also agreed to eliminate from the law a provision that would have allowed those with private insurance to choose a “public option,” that is, to choose a public plan similar to Medicare.

Republicans are strongly opposed to the ACA, and particularly to its insurance mandate. They believe it gives more power to the executive branch of government than the US Constitution allows, and they have brought the issue to the Supreme Court, which heard the case in March and will probably announce a decision in June. The ACA will be a central issue in the upcoming presidential campaign, with all potential Republican nominees pledged to seek immediate legislative annulment of the law if they are elected—assuming it is still standing after the Supreme Court hands down its opinion. The future shape of our healthcare system will therefore depend greatly on legal and electoral decisions made this year.

Some provisions in the ACA that have not yet been fully implemented, and a few initiatives already started by private insurers experiment with replacing the fee for service payment system with new arrangements. They pay providers for episodes of care, or encourage groups of physicians to share in any savings from assuming the comprehensive care of a designated panel of patients. These physician groups, affiliated with one or more hospitals, are called “accountable care organisations” and 32 such organisations in various parts of the country have recently signed contracts to try this model. It is much too early to know whether these initiatives will save money and be acceptable to doctors and their patients—or even whether they can function as expected—but they are being hailed by the Obama administration and by those who believe that modest changes in the present insurance and payments systems can control costs and improve the quality of care. I believe this is unrealistic, because they may never be fully completed and because they do not address the basic problems with the current system.9

Major reform is needed

I am convinced that the impending national cost crisis and the bankruptcy of our healthcare system will not be averted unless there is much more drastic and systematic reform.7 We will have to replace all our insurance systems, public and private, with a single public plan that guarantees universal access to prepaid, comprehensive care, and this plan should be funded by a progressive healthcare tax that all citizens (including government officials) must pay according to their means. Prepaid comprehensive care funded in this way would give government firm control over its total health expenditures while leaving decisions on the specific use of available resources where it should be—in the hands of physicians and their patients. By setting the rate of the healthcare tax, government would in effect be capping its costs. Any medical services provided outside the system would be at the patients’ expense.

To provide this kind of prepaid comprehensive care, we will need a reorganised medical care system based on private, non-profit, multispecialty group practices, in which physicians are paid largely or entirely by salary. If fee for service payment is to be largely or entirely replaced by prepaid, capitated payment, we will need medical organisations that can accept such payment and distribute an agreed, total percentage to physicians through salaries. The kind of multispecialty group practice that I have described in more detail elsewhere would be the ideal organisation to serve this function.10

There are ample reasons and considerable empirical evidence to suggest that group practices can deliver care more efficiently than unorganised physicians in solo or small, single specialty partnership practices who compete for income and depend on fee for service payment.11 Multispecialty groups usually include adequate numbers of primary care physicians who integrate and moderate the procedure based behaviour of the specialists. Most experts agree that substantial savings, as well as improved care, can be anticipated when primary care physicians collaborate with specialists in well organised groups. Without the fee for service incentive, prepaid salaried groups of this kind are less driven to unnecessary or duplicative services. And if the system is funded entirely by government without involving bills, the costs of insurance overheads and fraudulent billing are avoided. The potential savings would be enormous—a reduction of 30% to 40% in total expenditures would be a conservative estimate.

Multispecialty practices—the future

Multispecialty, physician managed group practices already exist in many parts of the US, mostly in the west, and least in the southeastern states. Some are well known, long established, and highly reputed. Only a few groups pay full salaries and most are still dependent on fee for service payment from multiple public and private insurers. However, the number of group practices in the country of all kinds and the number of physicians employed in these practices are increasing rapidly. There are at present over 400 such groups, employing nearly 200 000 physicians (about 25% of all practising physicians). These numbers are increasing at about 10% per year, according to information recently given to me by the American Medical Group Association, the organisation that represents most multispecialty groups.

There are many reasons why so many US physicians seem to be abandoning private practice in favour of seeking employment in large groups. Many starting physicians have acquired large personal debts during the course of their education and cannot afford the investment and financial risk of setting up their own practice. The supporting staff and facilities available in group practices and the assistance they offer with the business and administrative expenses of practice are very attractive; so are the retirement and fringe benefits. Furthermore, young US physicians nowadays are much more interested than formerly in practice settings that allow them to share responsibilities with colleagues and to work specified hours rather than assuming the heavy time demands of solo or small partnership practice. This is particularly true of women physicians, who shortly will constitute half of the physician workforce. But physicians of both genders are increasingly interested in a professional career that allows them more time for family and personal affairs. Employment (full or part time) in large groups provides that opportunity.

For their part, hospitals and physician organisations are eager to establish group practices because of the growing interest of public and private insurance plans in controlling costs by contractual arrangements with groups of physicians rather than individual practitioners. If this trend continues, soon the great majority of US physicians will be employed in large multispecialty groups and the prospects for major reform of the healthcare system will have greatly improved.

With so many physicians employed in multispecialty practices it would be much easier to institute new payment methods that replace insurance based reimbursement for itemised services with tax supported prepaid access to comprehensive care. If group practice proves popular with patients and physicians, and is increasingly perceived as an effective way to control costs and improve quality of care, it would also be much easier to persuade physicians in the groups to accept payment by salaries instead of fee for service. Popular pressure on Congress to enact reforms that are now politically impossible is more likely to be successful when most physicians are happily employed in multispecialty groups, and these groups are delivering good, affordable care.

Fierce resistance to reform can be expected from the private insurance industry, from Republicans in Congress, and from all those with vested financial interests in the status quo, so it is quite possible that the necessary legislation would be stalled until our health system faces bankruptcy, or even for an indefinite time after that. The misguided view that healthcare is primarily a business and that it requires the intervention of private insurance plans is so deeply embedded in US culture that legislation changing it to a universal right supported by government may be a long time in coming, and will certainly depend on how the political climate changes. But sooner or later reality will prevail, because a health system largely shaped by free market forces and heavily dependent on private insurance will never provide the whole US population with good medical care at an affordable cost.

The inevitable change in the health system, when it finally comes, won’t be achieved by a single legislative act of Congress. There will probably be stepwise increments—perhaps starting with reforms in various states. But I am convinced that tax supported, not-for-profit multispecialty group practice is necessary for the rescue of the US healthcare system. So the prospects for major health reform in the US depend on whether its physicians continue to choose employment in groups, and whether insurance plans are ultimately replaced by a single public payer that provides universal access to comprehensive care rather than itemised reimbursement for services.

Arnold Relman biography

Arnold Relman was editor of the New England Journal of Medicine from 1977 to 1991. He is the author of A Second Opinion: Rescuing America’s Health Care. He qualified in medicine at Columbia University, New York, in 1946, and then began research in nephrology and electrolyte balance while at Yale. In 1966 and 1974 he coedited two volumes of Controversy in Internal Medicine.


Cite this as: BMJ 2012;344:e3052


  • Competing interests: The author has completed the ICMJE uniform disclosure form at (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years, no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned, externally peer reviewed.