What is the cause of excess costs in US healthcare?

The question has come up again and again in our discussions on health care in the US and around the world, why does it cost so much more in the US when we get so much less?

The drug companies and their lobbyists are already out in force trying to make sure their pocketbooks aren’t hit by the inevitable reforms that are coming. In particular they insist drugs aren’t the problem in the US, it’s administrative costs! I would tend to disagree.

Based upon my experience working in the sytem, the main causes of excess costs I would hypothesize are the following (in order of importance):

  1. An excess of cost in administration far out of line with most countries around the world.
  2. Pharmaceutical costs – especially due to the effects of direct to consumer advertising (DTCA) encouraging use of more expensive, newer drugs (which is only allowed in the US), Medicare part D which forbids collective bargaining for lower drug prices, and a broken patent system that allows drug makers to patent and charge more for non-novel medications.
  3. The absence of a universal system that prevents risk-sharing, and causes the uninsured to avoid treatment until problems are more critical, and more expensive.
  4. Excessive reimbursement of physicians for procedural skills, rather than cost-saving physician roles such as primary care and family practice that emphasize early diagnosis and proper management of disease.
  5. Excesses of cost caused by “defensive medicine”. While torts themselves don’t cause a great deal of monetary damage, the culture they create is one of paranoia in physicians who make decisions with lawsuits in mind, rather than the interests of patients and society
  6. The excessive costs in ICU care, especially at the end of life, which may also be reduced by better EMRs with recording of living wills, and public information campaigns designed to inform people about the pain, invasiveness and futility of “doing everything” in the elderly.
  7. The absence of an electronic medical record that is universal which causes redundancy in testing as patients see new doctors who then order redundant tests because sharing of information is so inefficient.

But these may just be my biased views based on my own limited experience. Let’s see what the data show. The McKinsey Global Institute has generated a report on this, and has broken down the data according to the individual costs in our system, while comparing it to that of other countries.

The report available here is an up-to-date assessment of US health care costs and looks at many of the factors that I believe are involved in our excess costs. It is extremely detailed and goes into far many more topics than can be covered here, but we’ll hit the high points.

Now, let’s talk about why the US is so far above any other country in terms of our health spending:


First on my list, excess costs of administration. The McKinsey report has this figure showing where the excess administrative costs appear to be in our system.
i-2b5e31cf2acaf62b8c3b73b6aa6a8503-Drug Prices2.jpg

i-51b608f51abbcacdb24f53e612e59d77-Administration costs.jpg

The McKinsey report estimates these costs at 91 billion more than would be expected. The US spends ~16% of our 14 trillion GDP on healthcare, or about 2.25 trillion dollars. So 91 billion would account for about 4% of total spending, and if we should be spending about half as much per capita, maybe about 8% of the total waste is in this category. The McKinsey report, using a different calculation, described waste in administrative costs as 14% of our health care excess.

Costs in medicare administration have begun to shoot up too, concurrent with Bush administration efforts to subcontract administration to the private sector – an act despised by many physicians who find their non-reimbursement tactics to be dishonest scams. Administrative costs from 2003-2006 more than doubled, largely due to these practices.


Medicare used to be one of the most efficiently-administered systems in government until the Bush administration got its hands on it. By subcontracting administrative work to private contractors, suddenly 30% of the costs just go to profit, and payments to physicians for reimbursement have decreased accordingly. Most physicians I know under this regime have complained that they create barriers to registration, barriers to reimbursement, have very little recourse for rejected reimbursement, and then finally when everyone gets fed up with one of these subcontractors the government jumps to a new one. Then you have to start the process over again, learn the quirks of the new company, start a complaint file with them, and by the time you’re used to them, a new company gets the contract. This is the lie of the private system being better than government no matter what. I’m sure private contractors can do something better than public servants, but it certainly isn’t healthcare administration, and the data shows it, as costs ramped exponentially under this system despite relatively stable numbers of medicare subscribers.

The argument for public administration of health care is strengthened by these numbers, and based upon historically medicare administration costing only about 10% as much as private insurers’ administrative costs. One may also look at other public systems like the VA, in which administrative costs are a fraction of those with private insurers. Or, one may point out that in health care systems with public and private payers in competition, or highly regulated private payers, these costs are also substantially reduced, as should be clear from the first figure – most the comparison countries have private insurers, they are just better regulated.

Onto the second offender, pharmaceutical prices. The McKinsey report found that in the US we pay twice as much for drugs

i-1ba803d92583631acdafba7d876fa3ad-Drug Prices.jpg

Why? A few reasons. One is, we can afford to pay more so the drug companies know they can charge us more in the US. Thus, we subsidize health care around the world by paying higher for pharmaceuticals than other countries have bargained for. Another is direct to consumer advertising, which drives consumers to use drugs unnecessarily, and to use more expensive and newer drugs rather than generics. Finally, the largest problem is that our government, hijacked by these companies during the last legislation over health care, was forbidden from collectively bargaining for lower drug prices under medicare part D. A decision no other country in the world would be so stupid as to make as they didn’t have a political parties hugely corrupted by pharmaceutical interests like we do.

From the WSJ article, drug company execs are terrified at the prospect of collective bargaining.

Meanwhile, drug-industry executives worry that an overhaul of the health-care system could lead to too much government intervention. In addition to possibly establishing a government-sponsored insurance plan, lawmakers might give Medicare — the existing public program for the elderly and disabled — the authority to negotiate the prices for drugs dispensed through its Part D benefit. That could limit the prices pharmaceutical companies can charge.

Pharmaceutical executives argue that such steps would hamper drug makers’ ability to pay for costly research into new treatments. “It would knock our legs out,” says Lilly’s Dr. Lechleiter.

This, I’m sorry to say, is a bald-faced lie.

i-9cbd8347f558621ba666bc33b853cd40-novel drugs.jpg

Most drug company R&D is on the creation of sibling, or “me too” drugs that mimic existing mechanisms and enjoy equal patent protection to novel classes of drugs. Novel drug mechanisms are usually a product of the public R&D through institutions like the NIH. As long as our patent system rewards the creation of “me too” drugs, that is what we’re going to get and what drug companies will invest in. If we restructure the patent system to again encourage innovation, by shortening the patent protection of sibling drugs, or calling them what they usually are, generic equivalents, then we would not only have cheaper drugs but more novel drugs. See the work of Marcia Angell for more on this topic. When she examined the idea that drug company research produces new treatments what she instead found was the overwhelming majority (85%) of novel drugs resulted from government-funded research.

If we were to eliminate drug company R&D, the outcome would be less of these drugs that allow them to finagle their way past patent laws by making scibling drugs. If we eliminated direct to consumer advertising, their marketing budgets (far larger than their R&D) could instead go to research on drugs. If we reform the patent system so they no longer enjoy equal protection for non-novel substances, then maybe they’ll spend all of their R&D on new treatments for diseases rather than the shortcut to wealth that sibling drugs represent.

So, what is the overall cost of the pharmaceutical excess? The McKinsey report suggests about 98 billion dollars of excess costs comes from our use of more expensive drugs, non-generics, and our paying of higher prices for the same drugs relative to other countries. That would be about 4.5% of our total health care expenditure is on this excess, about 9% of waste according my simple half-is-waste calculation, or about 15% of total excess costs according to the McKinsey calculation.

My third hypothesis about excess cost was the “hidden tax” of lacking universal health care costs and an excess of cost from acute treatment of chronic disorders of the uninsured. An organization named Families USA has just made the news for their report on the hidden premiums and taxes created by the uninsured. As I’ve pointed out before, we already have a kind of universal health care, because doctors are ethically obligated to treat people whether they can pay or not. The government covers indigent patients, and the hospitals redistribute these costs onto other patients so they can run money-losing departments like ERs and mental health facilities. Families USA have estimated these costs of covering the uninsured to be $1000 on an average family’s health care premiums or about $368 for each individual. So, as you can see we are already “taxed” to cover some form of universal health care. Only this tax is non-transparent, it encourages poor use of health care resources like those of ERs and inpatient critical care, and it’s not justly distributed as it’s levied on all families independent of income. This cost represents tens of billions of dollars being spent invisibly, on inefficient care, and levied without a progressive structure. I’m not sure how much of it can be categorized as “waste”, but it is certainly a more wasteful way to provide care. The McKinsey report indicates outpatient care and especially same-day outpatient care (emergent visits) is the largest portion of the 650 billion dollars in excess spending (about two-thirds of this cost) and much of this is due to ER visits, imaging, and costs of acute care. The absence of universality, therefore, may be the most expensive source of waste in our system.

Fourth, we have the issue of physician compensation. In the US people can often chose to self-refer to specialists, who tend to be more expensive. We also have more specialists, more highly paid specialists, and more use of specialists.
i-efd0e03519bfb624d894c2624ac2acf6-physician cost.jpg
As you can see from the figure, specialists account for much of this excess cost. When studied, GP use tends to save money, because they manage care better, prevent critical problems from becoming worse, etc. Also the specialists tend to be procedurally-oriented, and make a great deal of money relative to time spend whenever they perform procedures. This is another area which should be addressed by reform. GPs, because they save money should be rewarded more for their time. Specialists, since they cost so much money, should probably have their reimbursement brought more in line with reality. Yes, I’m arguing against my future salary here. Overall, excess spending on physicians is about 60 billion dollars or about 2.7% of health care expenditure and 5.4% of waste based on my “half-waste” scheme. The McKinsey report doesn’t give a percentage of this factor, but based on their scale I suspect it’s under 10% of total waste. I would also ask people to give US physicians a little bit of slack on this topic, as for most physicians, medical education costs about 200-250 thousand dollars and generates huge amounts of debt. Other medical systems that educate physicians for free will have lower physician compensation at least in part due to the absence of student debt.

Last we have the factors of “defensive medicine”, end-of-life care, and absence of efficient medical records to prevent redundant care. These are going to be harder to quantify. For instance, take defensive medicine. If I had a patient who presented for, say, alcohol withdrawal, but it comes out in their history that two weeks ago they struck their head losing consciousness while drunk, they will likely end up in the CT scanner despite an absence of neurological findings. We could likely avoid such scanning if we had a conversation with the patient talking about the risks of avoiding the procedure, charting that the patient agreed to refuse the CT, and performing a thorough neurologic exam. However, I’ve found that in practice the tendency is to provide excessive care to avoid missing anything. Now, in such a case of an old head injury (and I’ve run into this many times) it is unlikely that even a positive CT-finding of an old intracerebral hemorrhage would change the care plan, but we’ll still do it, because it’s the kind of diligence required by defensive medicine. I’ve had this exact scenario occur, and in a patient with financial problems that a CT could put them into financial trouble being uninsured, but with just enough assets they would be expected to pay for their care. Examples of this behavior are rife in the system, and while I am generally opposed to tort reform, one can not deny that much of physician behavior is driven by fear of being sued. And while defensive medicine isn’t necessarily bad care (one could argue about the excessive use of radiological imaging as increasing cancer risks), it is certainly excessive care. From a societal perspective it could stand to be curtailed. Surveys of whether defensive medicine is practice indicate it is fairly prevalent especially among high-risk specialties.

Nearly all (93%) reported practicing defensive medicine. “Assurance behavior” such as ordering tests, performing diagnostic procedures, and referring patients for consultation, was very common (92%). Among practitioners of defensive medicine who detailed their most recent defensive act, 43% reported using imaging technology in clinically unnecessary circumstances. Avoidance of procedures and patients that were perceived to elevate the probability of litigation was also widespread. Forty-two percent of respondents reported that they had taken steps to restrict their practice in the previous 3 years, including eliminating procedures prone to complications, such as trauma surgery, and avoiding patients who had complex medical problems or were perceived as litigious. Defensive practice correlated strongly with respondents’ lack of confidence in their liability insurance and perceived burden of insurance premiums.

I can not find good data on the cost of such behaviors, and I’m concerned there is considerable ideological agendas influencing much of the writing (if someone has a good study please share it). Tort reform should not be about making physicians immune to societal correction, but at the same time physicians are unfairly penalized for expected complications and the adversarial system does not justly compensate those who have suffered from such complications. Tort reform should therefore emphasize “victim compensation” which would hopefully eliminate the need for physicians to practice cover-your-ass medicine and avoid adversarial relationships between physicians and patients who have had bad outcomes. Physicians should want their patients with bad outcomes to benefit from compensation, and receive additional care and benefits needed to address medical complications, but because lawsuits are the mechanism, physicians and patients instead end up in adversarial and hostile relationships when complications occur. This also negatively affects physician compensation as insurance premiums are quite expensive and an unfortunately large portion of your bill is going to yet more insurance premiums. Worse, in the tort system, patients and doctors both lose, only the lawyers win.

End-of-life care has long been recognized as devastatingly expensive, often futile and worse, my experience with it makes me concerned that patients’ families often don’t understand how invasive and painful ICU care can be. Their natural desire to want to “do everything” for their loved ones does not seem to be adequately balanced with an understanding of the invasiveness and unpleasentness of ICU care and some 22% of all deaths occur in the ICU. Bodily functions are one-by-one replaced by machines from ventilators in the throat, tubes into the stomach for feeding, dialysis for failing kidneys, and urinary and rectal catheterization for waste control. Blood is drawn daily, sometimes multiple times a day, not to mention finger sticks. Central lines are placed into major veins. Sepsis almost invariably ensues in these fragile patients and nosocomial infection is a constant threat. The more experience I have with ICUs the more I have become convinced that the ICU is not the place to die. But how do we know when medical care has become futile? How do we know when the ICU won’t lead to a miraculous turn-around? The answer is that we don’t. And while exceptions are rare, there are those elderly patients who walk out of the ICU (or are transferred to the floor) all the better for the aggressive interventions that critical care entails. However, locations that have reduced intervention towards the end of life appear to reduce costs from use of doctors and ICUs without a loss of quality of life.

It has been estimated that much of your total lifetime healthcare expenditure will occur in the last year of life. Some 30% of yearly medicare costs go to patients in their last year of life according to the Dartmouth Atlas Study but there is a great amount of regional variation in this spending. And without a crystal ball to give us data on when people are going to die, the default is to treat an 80-year-old in the ICU the same as we would a 20-year-old, despite the vanishingly small chance they will benefit with advancing age. We simply lack the data to have a good idea about who will benefit from ICU treatment and who will just have their inevitable death extended and made more expensive, undignified and painful.

Will the electronic medical record (EMR) come to the rescue? I think it might, and there are reports that having discussions with your doctor prevents aggressive end-of-life care without actually shortening life-span or worsening quality of life. In fact, aggressive care appears to result in worse quality of life. Having living wills and advance directives part of every medical record could go a long way towards less invasive and less expensive care end of life. Much of the redundancy in care we have is that every hospital has a different record system, and most primary care providers don’t have electronic records. If someone at an outside hospital gets a CT and is then transferred to another, more acute-care hospital, we at the level 1 trauma centers often have to repeat tests because we can’t wait for the data transfer which may take a day. That’s extra CT scans, extra MRIs, extra blood tests, and repeats of invasive and unpleasant testing. Further the EMR allows us to learn more about our patient populations and collect data which can be mined to figure out things like the probabilities of survival in an ICU. One of the unexpected benefits of the universal systems in other countries has been access to incredible data pools which allow us to more scientifically measure and administer care. It is clear though, without it, waste, inefficiency, redundancy and other problems ensue.

So, to sum up, the major cost drivers in our system stem from inefficient use. Possibly the most expensive are same-day and outpatient use of services which may be a reflection of our absence of universal coverage and acute treatment of problems better served by cost-saving GPs. Pharmaceutical costs come next with US citizens paying twice as much for drugs as our counterparts in other countries, largely due to our use of more expensive pharmaceuticals and our inability to collectively bargain with pharmaceutical companies. Administrative costs are also expensive, and privatization has increased costs even in the historically-efficient medicare system, these services need to either be centralized or regulated to prevent excessive profit-taking, and inefficiency. Physician compensation is higher in our country than in others, and may benefit from victim compensation funds, increased use of cost-saving GPs rather than specialist care, government subsidy of medical education, and reductions in specialist salaries through more even payment structure for procedures compared to office visits. The unknowns of defensive medicine likely contribute to our overuse of testing and imaging services as we also pay more than other countries for diagnostics, and I suspect this is a large portion of the cause. Finally, we could stand to address systemic inefficiencies through better information technology and the EMR which would reduce inefficiencies and redundant care as well as reduce the expense of end-of-life care through better data collection and collection of final directives and living wills.