With all the shouting going on about America’s health care crisis, many are probably finding it difficult to concentrate, much less understand the cause of the problems confronting us. I find myself dismayed at the tone of the discussion (though I understand it—people are scared) as well as bemused that anyone would presume themselves sufficiently qualified to know how to best improve our health care system simply because they’ve encountered it, when people who’ve spent entire careers studying it (and I don’t mean politicians) aren’t sure what to do themselves.
Albert Einstein is reputed to have said that if he had an hour to save the world he’d spend 55 minutes defining the problem and only 5 minutes solving it. Our health care system is far more complex than most who are offering solutions admit or recognize, and unless we focus most of our efforts on defining its problems and thoroughly understanding their causes, any changes we make are just likely to make them worse as they are better.
Though I’ve worked in the American health care system as a physician since 1992 and have seven year’s worth of experience as an administrative director of primary care, I don’t consider myself qualified to thoroughly evaluate the viability of most of the suggestions I’ve heard for improving our health care system. I do think, however, I can at least contribute to the discussion by describing some of its troubles, taking reasonable guesses at their causes, and outlining some general principles that should be applied in attempting to solve them.
THE PROBLEM OF COST
No one disputes that health care spending in the U.S. has been rising dramatically. According to the Centers for Medicare and Medicaid Services (CMS), health care spending is projected to reach $8,160 per person per year by the end of 2009 compared to the $356 per person per year it was in 1970. This increase occurred roughly 2.4% faster than the increase in GDP over the same period. Though GDP varies from year-to-year and is therefore an imperfect way to assess a rise in health care costs in comparison to other expenditures from one year to the next, we can still conclude from this data that over the last 40 years the percentage of our national income (personal, business, and governmental) we’ve spent on health care has been rising.
Despite what most assume, this may or may not be bad. It all depends on two things: the reasons why spending on health care has been increasing relative to our GDP and how much value we’ve been getting for each dollar we spend.
WHY HAS HEALTH CARE BECOME SO COSTLY?
This is a harder question to answer than many would believe. The rise in the cost of health care (on average 8.1% per year from 1970 to 2009, calculated from the data above) has exceeded the rise in inflation (4.4% on average over that same period), so we can’t attribute the increased cost to inflation alone. Health care expenditures are known to be closely associated with a country’s GDP (the wealthier the nation, the more it spends on health care), yet even in this the United States remains an outlier (figure 3).
Is it because of spending on health care for people over the age of 75 (five times what we spend on people between the ages of 25 and 34)? In a word, no. Studies show this demographic trend explains only a small percentage of health expenditure growth.
Is it because of monstrous profits the health insurance companies are raking in? Probably not. It’s admittedly difficult to know for certain as not all insurance companies are publicly traded and therefore have balance sheets available for public review. But Aetna, one of the largest publicly traded health insurance companies in North America, reported a 2009 second quarter profit of $346.7 million, which, if projected out, predicts a yearly profit of around $1.3 billion from the approximately 19 million people they insure. If we assume their profit margin is average for their industry (even if untrue, it’s unlikely to be orders of magnitude different from the average), the total profit for all private health insurance companies in America, which insured 202 million people (2nd bullet point) in 2007, would come to approximately $13 billion per year. Total health care expenditures in 2007 were $2.2 trillion (see Table 1, page 3), which yields a private health care industry profit approximately 0.6% of total health care costs (though this analysis mixes data from different years, it can perhaps be permitted as the numbers aren’t likely different by any order of magnitude).
Is it because of health care fraud? Estimates of losses due to fraud range as high as 10% of all health care expenditures, but it’s hard to find hard data to back this up. Though some percentage of fraud almost certainly goes undetected, perhaps the best way to estimate how much money is lost due to fraud is by looking at how much the government actually recovers. In 2006, this was $2.2 billion, only 0.1% of $2.1 trillion (see Table 1, page 3) in total health care expenditures for that year.
Is it due to pharmaceutical costs? In 2006, total expenditures on prescription drugs was approximately $216 billion (see Table 2, page 4). Though this amounted to 10% of the $2.1 trillion (see Table 1, page 3) in total health care expenditures for that year and must therefore be considered significant, it still remains only a small percentage of total health care costs.
Is it from administrative costs? In 1999, total administrative costs were estimated to be $294 billion, a full 25% of the $1.2 trillion (Table 1) in total health care expenditures that year. This was a significant percentage in 1999 and it’s hard to imagine it’s shrunk to any significant degree since then.
In the end, though, what probably has contributed the greatest amount to the increase in health care spending in the U.S. are two things:
1. Technological innovation.
2. Overutilization of health care resources by both patients and health care providers themselves.
Technological innovation. Data that proves increasing health care costs are due mostly to technological innovation is surprisingly difficult to obtain, but estimates of the contribution to the rise in health care costs due to technological innovation range anywhere from 40% to 65% (Table 2, page 8). Though we mostly only have empirical data for this, several examples illustrate the principle. Heart attacks used to be treated with aspirin and prayer. Now they’re treated with drugs to control shock, pulmonary edema, and arrhythmias as well as thrombolytic therapy, cardiac catheterization with angioplasty or stenting, and coronary artery bypass grafting. You don’t have to be an economist to figure out which scenario ends up being more expensive. We may learn to perform these same procedures more cheaply over time (the same way we’ve figured out how to make computers cheaper) but as the cost per procedure decreases, the total amount spent on each procedure goes up because the number of procedures performed goes up. Laparoscopic cholecystectomy is 25% less than the price of an open cholecystectomy, but the rates of both have increased by 60%. As technological advances become more widely available they become more widely used, and one thing we’re great at doing in the United States is making technology available.
Overutilization of health care resources by both patients and health care providers themselves. We can easily define overutilization as the unnecessary consumption of health care resources. What’s not so easy is recognizing it. Every year from October through February the majority of patients who come into the Urgent Care Clinic at my hospital are, in my view, doing so unnecessarily. What are they coming in for? Colds. I can offer support, reassurance that nothing is seriously wrong, and advice about over-the-counter remedies—but none of these things will make them better faster (though I often am able to reduce their level of concern). Further, patients have a hard time believing the key to arriving at a correct diagnosis lies in history gathering and careful physical examination rather than technologically-based testing (not that the latter isn’t important—just less so than most patients believe). Just how much patient-driven overutilization costs the health care system is hard to pin down as we have mostly only anecdotal evidence as above.
Further, doctors often disagree among themselves about what constitutes unnecessary health care consumption. In his excellent article, “The Cost Conundrum,” Atul Gawande argues that regional variation in overutilization of health care resources by doctors best accounts for the regional variation in Medicare spending per person. He goes on to argue that if doctors could be motivated to rein in their overutilization in high-cost areas of the country, it would save Medicare enough money to keep it solvent for 50 years.
A reasonable approach. To get that to happen, however, we need to understand why doctors are overutilizing health care resources in the first place:
1. Judgment varies in cases where the medical literature is vague or unhelpful. When faced with diagnostic dilemmas or diseases for which standard treatments haven’t been established, a variation in practice invariably occurs. If a primary care doctor suspects her patient has an ulcer, does she treat herself empirically or refer to a gastroenterologist for an endoscopy? If certain “red flag” symptoms are present, most doctors would refer. If not, some would and some wouldn’t depending on their training and the intangible exercise of judgment.
2. Inexperience or poor judgment. More experienced physicians tend to rely on histories and physicals more than less experienced physicians and consequently order fewer and less expensive tests. Studies suggest primary care physicians spend less money on tests and procedures than their sub-specialty colleagues but obtain similar and sometimes even better outcomes.
3. Fear of being sued. This is especially common in Emergency Room settings, but extends to almost every area of medicine.
4. Patients tend to demand more testing rather than less. As noted above. And physicians often have difficulty refusing patient requests for many reasons (eg, wanting to please them, fear of missing a diagnosis and being sued, etc).
5. In many settings, overutilization makes doctors more money. There exists no reliable incentive for doctors to limit their spending unless their pay is capitated or they’re receiving a straight salary.
Gawande’s article implies there exists some level of utilization of health care resources that’s optimal: use too little and you get mistakes and missed diagnoses; use too much and excess money gets spent without improving outcomes, paradoxically sometimes resulting in outcomes that are actually worse (likely as a result of complications from all the extra testing and treatments).
How then can we get doctors to employ uniformly good judgment to order the right number of tests and treatments for each patient—the “sweet spot”—in order to yield the best outcomes with the lowest risk of complications? Not easily. There is, fortunately or unfortunately, an art to good health care resource utilization. Some doctors are more gifted at it than others. Some are more diligent about keeping current. Some care more about their patients. An explosion of studies of medical tests and treatments has occurred in the last several decades to help guide doctors in choosing the most effective, safest, and even cheapest ways to practice medicine, but the diffusion of this evidence-based medicine is a tricky business. Just because beta blockers, for example, have been shown to improve survival after heart attacks doesn’t mean every physician knows it or provides them. Data clearly show many don’t. How information spreads from the medical literature into medical practice is a subject worthy of an entire post unto itself. Getting it to happen uniformly has proven extremely difficult.
In summary, then, most of the increase in spending on health care seems to have come from technological innovation coupled with its overuse by doctors working in systems that motivate them to practice more medicine rather than better medicine, as well as patients who demand the former thinking it yields the latter.
But even if we could snap our fingers and magically eliminate all overutilization today, health care in the U.S. would still remain among the most expensive in the world, requiring us to ask next—
WHAT VALUE ARE WE GETTING FOR THE DOLLARS WE SPEND?
According to an article in the New England Journal of Medicine titled The Burden of Health Care Costs for Working Families—Implications for Reform, growth in health care spending “can be defined as affordable as long as the rising percentage of income devoted to health care does not reduce standards of living. When absolute increases in income cannot keep up with absolute increases in health care spending, health care growth can be paid for only by sacrificing consumption of goods and services not related to health care.” When would this ever be an acceptable state of affairs? Only when the incremental cost of health care buys equal or greater incremental value. If, for example, you were told that in the near future you’d be spending 60% of your income on health care but that as a result you’d enjoy, say, a 30% chance of living to the age of 250, perhaps you’d judge that 60% a small price to pay.
This, it seems to me, is what the debate on health care spending really needs to be about. Certainly we should work on ways to eliminate overutilization. But the real question isn’t what absolute amount of money is too much to spend on health care. The real question is what are we getting for the money we spend and is it worth what we have to give up?
People alarmed by the notion that as health care costs increase policymakers may decide to ration health care don’t realize that we’re already rationing at least some of it. It just doesn’t appear as if we are because we’re rationing it on a first-come-first-serve basis—leaving it at least partially up to chance rather than to policy, which we’re uncomfortable defining and enforcing. Thus we don’t realize the reason our 90 year-old father in Illinois can’t have the liver he needs is because a 14 year-old girl in Alaska got in line first (or maybe our father was in line first and gets it while the 14 year-old girl doesn’t). Given that most of us remain uncomfortable with the notion of rationing health care based on criteria like age or utility to society, as technological innovation continues to drive up health care spending, we very well may at some point have to make critical judgments about which medical innovations are worth our entire society sacrificing access to other goods and services (unless we’re so foolish as to repeat the critical mistake of believing we can keep borrowing money forever without ever having to pay it back).
So what value are we getting? It varies. The risk of dying from a heart attack has declined by 66% since 1950 as a result of technological innovation. Because cardiovascular disease ranks as the number one cause of death in the U.S. this would seem to rank high on the scale of value as it benefits a huge proportion of the population in an important way. As a result of advances in pharmacology, we can now treat depression, anxiety, and even psychosis far better than anyone could have imagined even as recently as the mid-1980′s (when Prozac was first released). Clearly, then, some increases in health care costs have yielded enormous value we wouldn’t want to give up.
But how do we decide whether we’re getting good value from new innovations? Scientific studies must prove the innovation (whether a new test or treatment) actually provides clinically significant benefit (Aricept is a good example of a drug that works but doesn’t provide great clinical benefit—demented patients score higher on tests of cognitive ability while on it but probably aren’t significantly more functional or significantly better able to remember their children compared to when they’re not). But comparative effectiveness studies are extremely costly, take a long time to complete, and can never be perfectly applied to every individual patient, all of which means some health care provider always has to apply good medical judgment to every patient problem.
Who’s best positioned to judge the value to society of the benefit of an innovation—that is, to decide if an innovation’s benefit justifies its cost? I would argue the group that ultimately pays for it: the American public. How the public’s views could be reconciled and then effectively communicated to policy makers efficiently enough to affect actual policy, however, lies far beyond the scope of this post (and perhaps anyone’s imagination).
THE PROBLEM OF ACCESS
A significant proportion of the population is uninsured or underinsured, limiting or eliminating their access to health care. As a result, this group finds the path of least (and cheapest) resistance—emergency rooms—which has significantly impaired the ability of our nation’s ER physicians to actually render timely emergency care. In addition, surveys suggest a looming primary care physician shortage relative to the demand for their services. In my view, this imbalance between supply and demand explains most of the poor customer service patients face in our system every day: long wait times for doctors’ appointments, long wait times in doctors’ offices once their appointment day arrives, then short times spent with doctors inside exam rooms, followed by difficulty reaching their doctors in between office visits, and finally delays in getting test results. This imbalance would likely only partially be alleviated by less health care overutilization by patients.
GUIDELINES FOR SOLUTIONS
As Freaknomics authors Steven Levitt and Stephen Dubner state, “If morality represents how people would like the world to work, then economics represents how it actually does work.” Capitalism is based on the principle of enlightened self-interest, a system that creates incentives to yield behavior that benefits both suppliers and consumers and thus society as a whole. But when incentives get out of whack, people begin to behave in ways that continue to benefit them often at the expense of others or even at their own expense down the road. Whatever changes we make to our health care system (and there’s always more than one way to skin a cat), we must be sure to align incentives so that the behavior that results in each part of the system contributes to its sustainability rather than its ruin.
Here then is a summary of what I consider the best recommendations I’ve come across to address the problems I’ve outlined above:
1. Change the way insurance companies think about doing business. Insurance companies have the same goal as all other businesses: maximize profits. And if a health insurance company is publicly traded and in your 401k portfolio, you want them to maximize profits, too. Unfortunately, the best way for them to do this is to deny their services to the very customers who pay for them. It’s harder for them to spread risk (the function of any insurance company) relative to say, a car insurance company, because far more people make health insurance claims than car insurance claims. It would seem, therefore, from a consumer perspective, the private health insurance model is fundamentally flawed. We need to create a disincentive for health insurance companies to deny claims (or, conversely, an extra incentive for them to pay them). Allowing and encouraging aross-state insurance competition would at least partially engage free market forces to drive down insurance premiums as well as open up new markets to local insurance companies, benefiting both insurance consumers and providers. With their customers now armed with the all-important power to go elsewhere, health insurance companies might come to view the quality with which they actually provide service to their customers (ie, the paying out of claims) as a way to retain and grow their business. For this to work, monopolies or near-monopolies must be disbanded or at the very least discouraged. Even if it does work, however, government will probably still have to tighten regulation of the health insurance industry to ensure some of the heinous abuses that are going on now stop (for example, insurance companies shouldn’t be allowed to stratify consumers into sub-groups based on age and increase premiums based on an older group’s higher average risk of illness because healthy older consumers then end up being penalized for their age rather than their behaviors). Karl Denninger suggests some intriguing ideas in a post on his blog about requiring insurance companies to offer identical rates to businesses and individuals as well as creating a mandatory “open enrollment” period in which participants could only opt in or out of a plan on a yearly basis. This would prevent individuals from only buying insurance when they got sick, eliminating the adverse selection problem that’s driven insurance companies to deny payment for pre-existing conditions. I would add that, however reimbursement rates to health care providers are determined in the future (again, an entire post unto itself), all health insurance plans, whether private or public, must reimburse health care providers by an equal percentage to eliminate the existence of “good” and “bad” insurance that’s currently responsible for motivating hospitals and doctors to limit or even deny service to the poor and which may be responsible for the same thing occurring to the elderly in the future (Medicare reimburses only slightly better than Medicaid). Finally, regarding the idea of a “public option” insurance plan open to all, I worry that if it’s significantly cheaper than private options while providing near-equal benefits the entire country will rush to it en masse, driving private insurance companies out of business and forcing us all to subsidize one another’s health care with higher taxes and fewer choices; yet at the same time if the cost to the consumer of a “public option” remains comparable to private options, the very people it’s meant to help won’t be able to afford it.
2. Motivate the population to engage in healthier lifestyles that have been proven to prevent disease. Prevention of disease probably saves money, though some have argued that living longer increases the likelihood of developing diseases that wouldn’t have otherwise occurred, leading to the overall consumption of more health care dollars (though even if that’s true, those extra years of life would be judged by most valuable enough to justify the extra cost. After all, the whole purpose of health care is to improve the quality and quantity of life, not save society money. Let’s not put the cart before the horse). However, the idea of preventing a potentially bad outcome sometime in the future is only weakly motivating psychologically, explaining why so many people have so much trouble getting themselves to exercise, eat right, lose weight, stop smoking, etc. The idea of financially rewarding desirable behavior and/or financially punishing undesirable behavior is highly controversial. Though I worry this kind of strategy risks the enacting of policies that may impinge on basic freedoms if taken too far, I’m not against thinking creatively about how we could leverage stronger motivational forces to help people achieve health goals they themselves want to achieve. After all, most obese people want to lose weight. Most smokers want to quit. They might be more successful if they could find more powerful motivation.
3. Decrease overutilization of health care resources by doctors. I’m in agreement with Gawande that finding ways to get doctors to stop overutilizing health care resources is a worthy goal that will significantly rein in costs, that it will require a willingness to experiment, and that it will take time. Further, I agree that focusing only on who pays for our health care (whether the public or private sectors) will fail to address the issue adequately. But how exactly can we motivate doctors, whose pens are responsible for most of the money spent on health care in this country, to focus on what’s truly best for their patients? The idea that external bodies—whether insurance companies or government panels—could be used to set standards of care doctors must follow in order to control costs strikes me as ludicrous. Such bodies have neither the training nor overriding concern for patients’ welfare to be trusted to make those judgments. Why else do we have doctors if not to employ their expertise to apply nuanced approaches to complex situations? As long as they work in a system free of incentives that compete with their duty to their patients, they remain in the best position to make decisions about what tests and treatments are worth a given patient’s consideration, as long as they’re careful to avoid overconfident paternalism (refusing to obtain a head CT for a headache might be overconfidently paternalistic; refusing to offer chemotherapy for a cold isn’t). So perhaps we should eliminate any financial incentive doctors have to care about anything but their patients’ welfare, meaning doctors’ salaries should be disconnected from the number of surgeries they perform and the number of tests they order, and should instead be set by market forces. This model already exists in academic health care centers and hasn’t seemed to promote shoddy care when doctors feel they’re being paid fairly. Doctors need to earn a good living to compensate for the years of training and massive amounts of debt they amass, but no financial incentive for practicing more medicine should be allowed to attach itself to that good living.
4. Decrease overutilization of health care resources by patients. This, it seems to me, requires at least three interventions:
* Making available the right resources for the right problems (so that patients aren’t going to the ER for colds, for example, but rather to their primary care physicians). This would require hitting the “sweet spot” with respect to the number of primary care physicians, best at front-line gatekeeping, not of health care spending as in the old HMO model, but of triage and treatment. It would also require a recalculating of reimbursement levels for primary care services relative to specialty services to encourage more medical students to go into primary care (the reverse of the alarming trend we’ve been seeing for the last decade).
* A massive effort to increase the health literacy of the general public to improve its ability to triage its own complaints (so patients don’t actually go anywhere for colds or demand MRIs of their backs when their trusted physicians tells them it’s just a strain). This might be best accomplished through a series of educational programs (though given that no one in the private sector has an incentive to fund such programs, it might actually be one of the few things the government should—we’d just need to study and compare different educational programs and methods to see which, if any, reduce unnecessary patient utilization without worsening outcomes and result in more health care savings than they cost).
* Redesigning insurance plans to make patients in some way more financially liable for their health care choices. We can’t have people going bankrupt due to illness, nor do we want people to underutilize health care resources (avoiding the ER when they have chest pain, for example), but neither can we continue to support a system in which patients are actually motivated to overutilize resources, as the current “pre-pay for everything” model does.