Some would say that doctors, once licensed, aren't subject to lots of traditional regulation. A review of the Federal Register1 would in fact show that doctors face few explicit regulations governing:
• Patients they will and will not treat
• Therapies and protocols they use
• Where to administer treatment to their patients
• Whether and how to measure the results of their work
In many ways, however, doctors' choices on each of these four dimensions of practice are microscopically regulated—through the way they are paid for their services. Reimbursement has become the primary mechanism through which the regulation of doctors occurs in the United States.2 To the extent that doctors cannot afford to do things they are not paid to do, and will gladly do more of those things they are paid handsomely to do, the decisions about whether, when, and how much to pay doctors for the various things they do has unwittingly become one of the most pervasive and powerful regulatory mechanisms ever devised.
Most discussions about reforming health care run into a dead end when the participants realize that the regulatory system that we call reimbursement will not allow it. The reimbursement system is structured to sustain the status quo. Caregivers who do things the way they've always been done, or who make improvements within the present architecture of care, can get paid for what they do. Those who wish to disrupt the system by changing the very architecture of care, however, often are stymied by the specter that there literally is no money to be made from doing it. This is because disruptive innovations, being new to the world, just don't fit within the existing categories of products for which prices have been set and approved for reimbursement.
Health-care reformers have made compelling cases that improving the value of health care can't happen unless those who receive health services know what they cost and bear at least a share of the cost burden.3 At the same time, however, the belief that employers or the government are morally obligated to cover health-care costs has become a tenet accepted with near-religious fervor by most people in modern societies. We therefore seem bound in the tautologically tight paradox that employers, Medicaid, or Medicare must cover most health-care costs in ways that insulate providers and patients from the very market pressures that would normally force efficiencies, greater accountability, and the delivery of increased value.
What makes the encumbrance of reimbursement even more distortive and binding is that most prices insurers pay are not set by market forces. Rather, they are administered prices that reek of the pricing algorithms and backroom negotiations used in communist systems. Those who set or approve prices for medical products and procedures are typically physicians, health economists, and actuaries who are impaneled by Medicare and private insurers to tell them what they should pay.4
Not surprisingly, we reap the same inefficient results that characterized Communism. Hospitals aggressively pursue some types of procedures—like coronary bypass surgery, for example—that are highly profitable.5 And they often shun money-losing services such as psychiatric and trauma care, as well as services like preventive and primary care, which could save costs in the long run.6 Eventually, more and more people lose "access," as the services that aren't paid well or must be provided at a loss become harder to find. But the profits and losses aren't a reflection of value to their customers, society, or the forces of supply, demand, and competition: they are the phantom result of inaccurately set prices that are grossly out of line with costs. Yet as in the communist system, we muddle along because the prices that are set mistakenly high roughly offset those set mistakenly low, allowing most hospitals and physicians' practices to eke out a modest profit after all cross-subsidization is complete.
The purpose of this chapter is to define the changes that need to be made in reimbursement systems so they can efficiently facilitate disruptive innovation, and to describe how policy makers, insurers, and employers can successfully implement these changes. We'll do this in six sections. We begin by recounting the history of health insurance and reimbursement, to give our readers a sense of how we got to where we are today. Second, we'll review the systems commonly used throughout the world. Then, in the third section, we'll explore more deeply the problems that these reimbursement systems have created in the health-care industry. In the fourth and fifth sections we'll offer our recommendations for two reimbursement systems—integrated capitation and a pairing of high-deductible insurance and health savings accounts—that can overcome the problems created by existing products. The final section will offer additional suggestions for helping the uninsured poor access the right kind of health care.
Before the advent of modern medicine in the early 1900s, the cost of caring for serious illnesses wasn't financially devastating. It was affordable because family members cared for the patients at home, or charitable organizations brought them into community facilities. Because there wasn't much that doctors and hospitals could offer the gravely ill, life insurance—not health insurance—matched the reality of medicine. Health insurance policies that were sold in that era resembled what we now call disability insurance—protection against lost wages due to illness or serious injury rather than against the costs of medical care.
By the 1920s to 1940s, however, general hospitals had become capable enough that patients could recover from some diseases that were previously fatal—and this care was expensive enough to be financially devastating. In an era when families struggled to save five dollars per month, the average hospital bill of $140 in 1928 could send a family to financial ruin.7 A health insurance industry therefore emerged to help people hedge against the costly event of catastrophic illness or accident.
During the Great Depression, a set of pioneering not-for-profit companies that became known as Blue Cross began selling prepaid insurance plans that guaranteed hospital services in exchange for a low annual fee. Another set of companies that became Blue Shield sold similar plans covering physician services.8 Regulators forced the companies of Blue Cross and Blue Shield, because they were not-for-profit, to price by community rating, regardless of individuals' health condition. All people in a community were therefore charged the same price for the same coverage.
The health insurance market expanded rapidly when for-profit insurance companies got into the act. Not facing the same regulatory restrictions as the Blues and similar state-approved insurers of last resort, for-profit insurers were able to offer experience-rated plans, enabling them to "cherry pick" lower-cost younger and healthier customers with lower prices. This new type of health insurance was sold by individual agents, mirroring how home and life insurance were sold. The cost of all health insurance was paid by individuals. As reflected in Figure 7.1, the portion of people covered by these plans grew from less than 10 percent in 1940 to nearly 80 percent by the 1970s.9
FIGURE 7.1 Trends in proportion of population covered by different types of insurance
Even while this self-paid market for health insurance was in its infancy, a few industrialists in the 1930s, including Henry Kaiser in California, began covering the cost of health insurance as an employee benefit. Kaiser's employee health initiatives began in the Mojave Desert with the construction of the Colorado River Aqueduct during the Great Depression. Sidney Garfield, a physician who had begun treating Kaiser's construction workers at a small desert hospital, and Harold Hatch, an insurance agent, devised a new plan that didn't just insure against catastrophic costs, but involved prepayment for all medical care. At five cents per day, this was a bargain to Kaiser. It helped him recruit workers into that forbidding environment and to keep them healthy on the job.
Congress enacted pervasive price controls, in the Stabilization Act of 1942,10 to keep inflation under control amidst the massive budgetary deficits during World War II. This law limited wage increases but permitted companies to provide employee insurance plans as a means for recruiting and compensating workers. Most employers offered pension plans and life insurance. But Henry Kaiser, with fresh memories of the impact that covering employees' health costs had in building the Colorado River Aqueduct, used this Act to recruit needed employees to his Long Beach, California, shipyards. His strategy was not just to pay for employees' health-care costs, but to provide the care as well through company-operated clinics—in hopes that this would help keep employees healthy and out of the hospital. Dubbed "Permanente," this was one of the nation's first health maintenance organizations (HMOs).11 In 1944, Kaiser opened the Permanente Health Plan to the general public. It proved to be especially popular with union members, who were attracted by the affordability and comprehensiveness of coverage compared to the traditional insurance system.
After World War II, employers began to expand health insurance offerings into what became known as "major medical" plans, which covered a broader range of the costs of major illnesses or serious injuries that had not been covered by the basic hospital insurance plans introduced earlier. The first major medical policy was written by Liberty Mutual Insurance Company in 1949, and similar plans rapidly captured the private insurance market. By 1953, 1.2 million people (roughly 1 percent of all Americans) were covered by major medical plans.12
In the first complete overhaul of the federal income tax system since its establishment in 1913, the Internal Revenue Act of 1954 (P.L. 83-591) declared employer contributions to employee health plans to be deductible business expenses. Health insurance was now formally recognized as a tax-advantaged form of compensation, and employers began scrambling to offer health insurance in lieu of increasing wages. By 1962, 38.2 million people (25 percent of all workers) had employer-provided major medical insurance. This increased to 92.6 million by 1969, and, with the help of the HMO Act of 1973, to 117.3 million by 1977 (79 percent of all workers).13
To differentiate themselves in the markets for talent and for insurance, employers and health plans in the 1960s next began to promote "comprehensive" coverage, which included reimbursement for day-to-day health-care expenses. The best employers offered the most comprehensive plans—with the least amount of employee contribution—to attract and retain the best employees. Before long the two types of employer-provided health-care assistance—true insurance against catastrophic illness, and reimbursement for day-to-day health-care expenses—had been combined into single-package health plans. This unwittingly induced a pervasive sense of entitlement that today burdens employers who struggle to remain competitive in the global markets for their products. As we did in Chapter 6, we'll term this bundle of two very different products—insurance and reimbursement—as "health assistance."
The dominant form of assistance today is fee-for-service (FFS), which gives providers a clear path to revenue: the more services you render, the more you get paid. In America's system today, as a result, a lot of providers offer a lot of care. Studies concluding that up to half of all medical services performed in the United States are medically unnecessary pin much of the blame on the financial incentives embedded in fee-for-service.14
Medicare and Medicaid are government-sponsored programs for particular populations not covered by employer-offered insurance. These programs resulted from the deepening divide between coverage of the younger, healthier workforce and the indigent and elderly populations. The push for universal coverage had existed since the time of Roosevelt's New Deal, as the prospect of leaving the neediest and sickest patients without health care has long been unappealing to many Americans. Interest groups led by the American Medical Association, however, feared that a universal insurance system would harm the physician-patient relationship (which some would say was a smoke screen for harming their pocketbooks). They fought off attempts to create such an infrastructure until 1965, when, in a compromise, Medicare and Medicaid were established to provide coverage to the elderly and needy.15
Part of this compromise was a pledge to build these programs around fee-for-service payments. This seemingly innocuous decision has been a key driver of the unsustainable growth in spending over the following decades as health-care costs continued to rise and the number of elderly enrollees increased.
By the 1980s it became clear that medical costs had spiraled out of control. The fee-for-service model of reimbursing physicians and hospitals received much of the blame. Capitation was a mechanism designed to solve this problem. For a fixed fee set annually, Health Maintenance Organizations like Kaiser Permanente agreed to provide all the care that each of their covered patients needed.
Capitation continues to thrive today as a fruitful health assistance mechanism used in managerially integrated provider organizations such as Kaiser. It eliminates the incentive that fee-for-service reimbursement creates for providers to give more care than is needed, and ostensibly gives providers an incentive to engage in wellness care and preventive services to keep their patients healthy. Capitation actually encourages the development of disruptive business models within these integrated provider organizations, in that using lower-cost venues of care and lower-cost caregivers such as nurse practitioners and physician assistants drives greater surplus or profitability.16 We term this type of capitation, which is practiced within integrated provider organizations, as integrated capitation.
What caused capitation to come off the rails was when non-HMO insurers co-opted the concept by knitting together networks of independent primary care physicians, specialists, and hospitals to replicate the cost control mechanisms of HMOs, without limiting patients' choice of a personal physician. Primary care physicians were given stewardship over a fixed fee per capita (hence the rubric capitation), and they became "gatekeepers" who granted or denied access to more expensive levels of specialist and hospital care.
Nonintegrated capitation generally fell into disfavor for three reasons, the first of which is that because employees change jobs, and employers often change the health plans they offer employees, the average tenure of a person in a particular health plan through the 1980s and 1990s was only three years.17 Thus, it was not in the economic interest of insurers or providers to incur the cost of care that could prevent costlier, more serious problems that had an impact beyond the three year horizon. Rather, the incentive was to minimize spending on those diseases in the short term, because the statistical odds were that the bill for those patients' more serious problems would arise on someone else's watch—often that of Medicare.
The second problem was that capitation restricted freedom of choice. Some consumers were comfortable choosing a health-care system, such as an HMO, rather than choosing a doctor. But others had grown accustomed to the freedom to choose their own doctors and hospitals at every level. They bristled aggressively when their gatekeepers refused access to second opinions or to reputable specialists because they were deemed unnecessary or were outside the network. Because a key reason why employers began offering health coverage in the first place was to attract and retain the right employees, most employers decided that the potential savings weren't worth jeopardizing employee goodwill.
The third problem arose because of the nonintegrated structure in which capitation was used. Capitation works well in integrated systems like Kaiser. But gatekeeper physicians, as independent businesspeople, found themselves playing a zero-sum game with specialists, who were themselves independent business-people.18 If the gatekeepers referred a patient to a more expensive specialist, they made less money; if the gatekeepers tried to restrict referrals, the specialists complained that they were making less money. Neither had been equipped with the perspective of systemwide costs and benefits of the decisions they were asked to make. When they denied access to more expensive care, gate-keeper physicians faced acrimonious confrontations with their patients, with whom trusting relationships were important. And when they made access to expensive care too easy, gatekeeper physicians had their judgment questioned by the insurance carriers, who were in search of opportunities to squeeze down costs even further. Hence, many independent physicians as business owners came to dislike capitation intensely. Salaried physicians working in integrated provider organizations didn't feel this Catch-22 as much, because such referrals did not significantly affect their income.19
By 1998 a backlash by employees and providers in noninte-grated systems led most employers to return to more traditional fee-for-service plans, and health-care inflation resumed its rapid rise, returning once again to double digits in 2003. From this point on, the strategy of many employers has been to shift more of these costs to employees. But cost-shifting is not cost-reduction.
Today's methods of health assistance—which have reverted strongly back to fee-for-service for the reasons described above—create three major distortions to the efficacy and efficiency of health care. First, they preserve costly providers rather than enabling disruptive ones to emerge. Second, they dictate the price of services, and as a result create artificial bubbles of profitability and unprofitability in different sectors of the industry—thereby misdirecting the flow of investment in new products and services. And third, their contracting practices actually drive hospitals' costs up, not down. We'll consider each of these distortions below.
The first problem that the present assistance system creates is that fee-for-service traps health care in high-cost institutions, to the exclusion of disruptive models. There is no better way to illustrate this perversion than through the history of payment for dialysis patients.20
Dr. Willem Kolff created the first drum dialyzer in 1943.21 Its widespread use in the Korean War cut mortality from acute renal failure in half. After this success, hospitals quickly built inpatient dialysis units for their acute renal failure patients, to tide them over until the cause of temporary kidney failure could be addressed. A handful of hospitals opened wards for patients with chronic renal failure, but this was impractical because it essentially entailed permanent hospitalization.
This changed in 1960, when Belding Scribner developed the first arteriovenous, or AV, shunt—a Teflon-coated, U-shaped tube that offered long-term, convenient access to a patient's blood vessels for dialysis. Once the Scribner Shunt was inserted, no further surgery was necessary. While patients still needed dialysis for the rest of their lives (or at least until transplant), they could now come and go as they pleased between dialysis sessions. Recognizing the potential for his invention, Scribner founded the world's first outpatient dialysis center in 1962, and the field of nephrology grew tremendously to handle the burgeoning number of chronic renal failure patients whose life expectancies were suddenly extended through dialysis. In the language of Chapter 5, renal failure was transformed from an acute to a largely chronic disease.
Scribner's AV shunt enabled a truly disruptive business model. Today, there are 4,200 dialysis centers in the United States. Patients can travel far from home by scheduling treatments at dialysis centers around the country. Dialysis nurses, rather than nephrologists, and clinics, rather than hospitals, could provide most of the care. While dialysis continues to prolong the lives of 350,000 Americans today, the cost of care has been tremendous: $32.5 billion was spent on care for end-stage renal disease patients in 2004, and this amount is expected to grow at 9 to 10 percent each year.22
But the story does not end there. Just as outpatient dialysis centers disruptively drove hospitals from this market, in-home dialysis has tremendous potential for disrupting the dialysis centers. The first in-home machines of the 1960s weren't pretty—they were typically converted washing machines that required costly modifications to the home electrical and plumbing systems. The machinery was complex and difficult to operate. But as the technology for home dialysis improved, the added convenience and privacy it offered generated strong demand. By 1972, 40 percent of the 11,000 dialysis patients in the United States were on home hemodialysis.23 Today, machines like System One from Lawrence, Massachusetts–based NxStage are just the size of a microwave, enabling portability within the house and away from home. There is no need to modify electrical and plumbing systems. Patients follow simple rules to operate the device.24 Home hemodialysis is about 40 percent less costly than clinic-based dialysis.25 Adopting this disruptive technology would have saved Medicare $3.9 billion in 200526—chump change for some, but for those of us who work for a living, that's a lot of money. What's more, in-home dialysis typically is done daily, compared to the thrice-weekly regimen offered in clinics—better matching normal human physiology and very possibly leading to improved health outcomes.
Despite the advantages in cost and convenience that disruptive home hemodialysis offers, however, the market is moving away from it, not into it. Over the past 14 years, in-center hemodialysis has grown 7.25 percent annually.27 Only 0.6 percent of all patients—fewer than 2,000, compared to 11,000 in 1972, when the technology wasn't nearly as good—are on home hemodialysis today.28
What derailed the disruption? Fee-for-service reimbursement. In 1972, as renal care clinics were rolling out, Congress created the End-Stage Renal Disease (ESRD) program to guarantee fully reimbursed dialysis to anyone with kidney failure. ESRD remains the only medical condition that has ever been given a legislated guarantee of access to reimbursed care for everyone.
Because of their fixed cost investments, owners of dialysis clinics like Fresenius and DaVita profit by keeping their clinics full. Further, nephrologists often have a financial stake in the clinics in which they work, and are cited by patients as the most influential factor in determining whether they are dialyzed in the clinic or at home. They comprise a value network that has been very successful to date, having helped millions of patients with kidney failure. But can we expect them to urge patients toward home dialysis? Not when they profit so handsomely from guaranteed fees for high-cost service.
As evidence that reimbursement reform would facilitate disruptive business models in the absence of such fee-for-service, NxStage reports that 30 to 40 percent of its patients are private-pay.29 And in New Zealand, where consumers are much more engaged in determining when and how to start dialysis, over 25 percent of hemodialysis candidates are treated at home.30
How can we be so concerned with exploding health-care costs and be pouring fee-for-service fuel on the fire? The reason is that, given the present independent structure of hospitals and physicians' practices, there is no alternative. That's why separation of the different business models that are now conflated within hospitals and doctors' offices is such an important initial step, and why integrated entities—employers and integrated fixed-fee providers—need to emerge to wrap their arms around this problem. Then and only then will we discover that FFS is not the only way, let alone the best way, to transact business in health care. FFS will persist in solution shops. But we'll be able to bill, guarantee, and pay for work done in value-adding process businesses on a price-for-outcome basis, and gain the benefits of network services on a fee-for-membership or fee-for-transaction basis.
Dr. Robert Nesse of Mayo Clinic described the present condition with this query: "What would the cost of [a] hamburger at TGI Fridays be if, instead of paying for the outcome of good food delivered in a congenial location by friendly service, we actually just paid for the number of cooks . . . and how many wait staff that went by . . . What would happen to the price of a hamburger?"31
The second of the three involuntary evils created by today's health assistance industry is the system by which Medicare and insurance companies' formulas and appellate processes determine the prices of various health-care products and services. Medical fees were initially reimbursed by insurers based on "usual, customary, and reasonable" (UCR) charges determined by the hospitals and physicians that provided the care. As we described in Chapter 3, however, the conflated business models and the almost limitless number of pathways that patients can take through hospitals makes it literally impossible for the typical hospital to allocate its massive bucket of overhead costs to individual patients and procedures with any degree of accuracy. What's worse, the fees are based on costs, not the value created by a procedure. Except for the focused operators of value-adding process hospitals and clinics, neither the providers nor the insurers have an accurate sense of what real costs and real value are for most services offered.
As Medicare, Medicaid, and private health assistance companies pervasively inserted themselves between patients and providers, the market ultimately evolved toward what economists call monopsony—where a few huge, powerful buyers essentially determine the prices they will pay to their more fragmented suppliers. As market power shifted, health assistance firms essentially decided that while the charges from hospitals and physicians might be usual and customary, they almost certainly were not reasonable. They therefore devised formulas of the genre of the cost-based time-and-motion studies that characterized the work of Frederick Taylor in manufacturing a century ago.
The insurers' algorithms incorporated physician labor, indirect expenses, equipment costs, geographic location, and many other factors in calculating amounts for reimbursement. Because many of these formulas are based upon methods and technologies commonly used 20 years ago, the pace of change has rendered the resulting fee schedule hopelessly inaccurate. Although the fee schedule is constantly updated, having undergone thousands of corrections since its inception, the reimbursement payments it dictates often bear little relation to the value of the services rendered. As a result, health assistance companies and providers often find it easier to simply pass cost increases on to the employers, rather than argue with each other about prices.32 "They're spending my money as if it were my money," the CEO of a major manufacturing company lamented to us.
The result is that some medical services—particularly those that are procedure-based—have remained or become wildly profitable, while others, including office visits and out-of-office care, are undercompensated and therefore undesirable. Because profits attract investment, the de facto but pervasive economic regulation through this pricing mechanism has transformed America's health-care economy into a centrally planned one. The difference between a communist central planning system and this one, however, is that those who were pulling the strings in the Kremlin, for instance, were consciously empowered to do so. In America it is inadvertent central planning.
And it gets worse. In the present system of administered pricing, Medicare, Medicaid, and the health assistance companies simply cannot establish unique pricing algorithms for each nuanced improvement in drugs, devices, or services. Instead, they have created major groupings of these things, and they analyze and administer pricing by these categories—each of which is known by its Current Procedural Terminology (CPT) code under a system administered by the American Medical Association (AMA). A drug or device might have the disruptive potential to solve a problem in a cost effective but unique way, but the cost of getting a unique CPT code established for this new product is so high, and the probability that the AMA—the organization representing those who will be disrupted—will approve a disruptively positioned product is so low, that even the most determined disruptors can become disheartened. It is far easier instead to force-fit potentially disruptive innovations into existing CPT categories. That act, however, typically juxtaposes the disruptive technology head-on, on a sustaining basis, against the price and performance of established technologies in the existing market. In a world of fee-for-service and passed-through costs, the decision makers simply have no incentive to adopt disruptions.
The third way in which today's health assistance programs impede the efficiency of the health-care industry is through the practice of insurance companies negotiating volume discounts with hospitals through blanket contracting. This badly distorts the signal and amplifies the noise to providers and investors about where the opportunities for improving health care actually are. These discounts are an illusion—they are achieved by hospitals raising prices on patients who aren't covered by the blanket contracts with major insurers.
The competitive structure of most other industries causes companies to differentiate themselves—to find products or services that they can provide better than any competitors, and then to focus their resources on what they're best at providing. The process of winning contracts from the insurance companies, in contrast, compels general hospitals to offer the full breadth of services the insured population might need—to offer every service that every other hospital is offering. This exacerbates the complexity of general hospitals, driving up their over-head costs. It pits one-size-fits-all businesses into competition with other one-size-fits-all businesses, where every hospital aspires to be the best in everything. No manager would predict that such strategies will lead to efficiency and quality. Because disruptive care delivery institutions will be more focused, these contracting practices continue to funnel most patient volume to high-cost general hospital business models, preserving their prosperity despite their high costs, rather than directing patient volume to the disruptive providers.
Health assistance companies assert that they're creating value by bundling the volumes from multiple employers to negotiate lower prices from hospitals. But this sort of negotiation-for-discounts only creates lower costs if the suppliers are earning excess profit, or if suppliers' competitive markets are so lax that they have lost control of their costs. Neither of these situations characterizes today's hospitals. In a good year, the typical hospital's profit will be 2 percent of revenues.33 The high cost of hospitalization isn't driven by the excess profits of general hospitals. The costs are simply inherent to the one-size-fits-all value proposition they offer. Squeezing the prices paid to hospitals won't cause them to become markedly more efficient—any more than the corporate bean counters could squeeze more cost out of Noelle Allen's Pontiac plant at Michigan Manufacturing Corporation. The cost in hospitals is created by the conflation of multiple business models, and the complexity of service offerings within each of those models, all under one institutional roof. The contracting practices of health assistance companies exacerbate this practice and continue to drive costs up, not down.
In the remainder of this chapter we'll advocate that we need two types of payments products in tomorrow's health-care system. When employers begin directly managing employees' health care to ensure that they find providers whose business models best suit the nature of each problem, a combination product of high-deductible insurance with health savings accounts will help our system become better. And in cases where integrated entities are providing care, capitation is the best answer.
Insurance creates value by hedging against low-probability events that have financially devastating results. Insurance is an important component of our financial plans to build and protect our prosperity: it protects us and our families from the financial consequences of death, disability, fires, and accidents. Hedging against devastating hospital costs was the genesis of health insurance as well—it emerged in the 1920s when hospital care became a viable but unaffordable pathway to recovery. We call products like these "true insurance."
Ever since medical expenses began to rise dramatically in the 1920s, the need to insure against the cost of financially catastrophic illness has been remarkably unchanged. Figure 7.2 compares the distribution of medical costs across the U.S. population in 1928 and 2002. In both years, Pareto's law was apparent: 20 percent of the population accounted for 80 percent of health-care costs, and 5 percent of the population accounted for 50 percent of all costs. Yet because it is so difficult to predict exactly who will fall into that 5 percent, insurance is necessary for all but the very rich, to protect ourselves from this financial risk.
But insurance makes little sense for events that have a high probability of occurrence and recurrence and which are not financially devastating. One would not purchase insurance against the possibility of needing to pay for clothing and electricity, for example. To profitably cover the sales and administration costs of providing such insurance, premiums would need to be priced above the ongoing costs of those predictable events—meaning it is cheaper for people simply to pay those costs out of pocket rather than route the money through a third-party claims processor.
FIGURE 7.2 Distribution of medical costs: 1928 vs. 2002
When employers began to pay for low-priced, predictable, recurrent health-care events, it was a tax-advantaged form of employee compensation. When insurance companies lumped that benefit together with true insurance in "comprehensive health plans," however, they lumped a sensible, value-creating insurance product with a reimbursement service that actually destroys economic value because of its administrative overhead.34
Just as mutually incompatible business models within hospitals need to be teased apart in order to appropriately price the services of solution shops and value-adding process businesses, these mutually incompatible health insurance and reimbursement products need to be separated so they create rather than destroy economic value. This is the rationale behind the unbundled pairing of high-deductible insurance (HDI) and Health Savings Accounts (HSA) that more and more companies are offering their employees. Unbundling comprehensive health plans into these constituent parts is, in our view, one of the most important reforms to be made in health care. Where employers cannot or choose not to link their employees into an integrated health system that uses capitation and is aggressively implementing disruption, an HDI-HSA plan will be a necessary element of the new disruptive value network that major employers will need to orchestrate.
Health savings accounts were formally enabled by the Medicare Prescription Drug, Improvement, and Modernization Act signed by President George W. Bush in 2003. These accounts offer a tax-free, portable savings vehicle to help people pay for low-ticket, relatively predictable and recurrent medical expenses. As such, HSAs typically are offered with a high-deductible insurance product. Though the concepts underlying HSAs had already existed for at least 20 years, the Medicare Modernization Act relaxed a number of restrictions on eligibility for enrollees.
Briefly, here's how health savings accounts work. Our employer says to us, "It's costing us $10,000 per year to provide health assistance to your family—and you're contributing an additional $3,000 per year to your health-care costs in the form of copayments and other expenses we don't cover. From now on we're going to spend $5,000 per year to purchase a true umbrella insurance policy for you—a high-deductible health plan—to protect your assets and cover the cost of unpredictable high-cost medical events that arise. But we're going to put the other $5,000 into an account at a firm like Fidelity Investments.35 It will be linked to the 401(k) retirement account you already have there.
"Just like our expenditures to cover your health-care costs in the past, this amount that we deposit into your HSA account will be a before-tax expense to us. And just like you can make before-tax contributions to your 401(k) account, you can contribute additional money, before it is taxed, to your HSA to maximize your savings beyond the $5,000 we'll be putting in for you.36 You should invest this money just as you do in your 401(k), and the earnings will compound tax-free. In other words, this can be a huge enhancement to your retirement package.37
"Until you hit the level of spending on health-care costs at which the high-deductible insurance kicks in, you'll need to pay for those costs from your HSA. This means that the more frugally you manage your health-care costs, the more will accumulate, tax-free, in your HSA account. When you retire, this money and the money in your 401(k) are yours to use in any way you wish.
"And one more thing. Our costs of insuring your family have been increasing at about 10 percent per year. If we keep shelling money out for health-care costs, our company will go under—and none of us will have jobs. So we're not going to increase this HSA contribution beyond $5,000 per year. But you'll be okay. The typical mutual fund in which you're likely to invest your HSA balance historically appreciates at a compounded annual rate of about 8 to 10 percent—so you'll be able to keep up with the inflating costs of your health care.38 And if the disruptive business models that Christensen, Grossman, and Hwang recommended in their book come to fruition, you'll be much better off than you are today. You'll have better care for serious problems. Access to the day-to-day stuff will be much more convenient. And you'll have a significantly larger retirement nest egg."
Figure 7.3 diagrams how health savings accounts will fit with high-deductible insurance. HSAs will be used for lower-cost, recurrent health-care expenses. Insurance will kick in after an annual deductible has been incurred. Whether there is a gap between the amount that could be covered by the HSA and the point at which insurance would kick in will vary by employers' plans and the propensity of individuals to economize their health-related spending.
FIGURE 7.3 Relative roles of health savings accounts and high-deductible insurance
In short, HSAs are meant to encourage users to spend wisely by giving them control of their health-care dollars. It removes the drawbacks and restraints of the reimbursement system and gives price-sensitive consumers the motivation to seek value as they have defined value for themselves. At the same time, the pursuit of health also becomes a mechanism for increasing wealth. Individuals who practice healthy behaviors will generally see their long-term savings increase by a greater margin.
The jobs-to-be-done model that we presented in Chapter 1 (remember the milkshake?) helps us evaluate the strengths and weaknesses of various forms of health assistance. Two specific implications of the model merit special attention.
First, the process of paying for health care isn't a job to be done—it is an experience that customers go through as they hire a provider to do the fundamental jobs of becoming or remaining healthy. More convenient methods of payment will typically be preferred over less convenient ones. Similarly, cost isn't a job either. It's a characteristic of each of the competing products and services that might be hired to do the job.
And second, if we want to reform the present reimbursement system, the reform has to get the job done for employers, providers, politicians, patients, and insurers. Each has veto power. If one presents an insurmountable hurdle, then that group needs to be cut out of the new system—or the reform will sputter and fail.
While we would welcome others' attempts to delve deeply into this question, we suggest that among patients, providers, suppliers (such as pharmaceutical, biotechnology, and device companies), politicians, employers, and health assistance companies, there are eight jobs that need to be done that impact insurance and reimbursement. We'll describe each briefly here, and then summarize in Figure 7.4 how well various insurance and reimbursement products do these particular jobs.39
1. Help me and my family to become healthy. Almost universally, becoming well gets top priority when people become sick.
2. Help us to remain healthy. One needs only to stand along running trails, or study sales figures for natural foods and nutriceutical products, to see that this is an important job that many (but not all) people are trying to do. Many of those with illnesses like obesity, diabetes, heart disease, asthma, and nicotine addiction either don't feel the urgent need to do this job on a daily basis or find the challenge simply too daunting to overcome.
3. Help us achieve financial ability. Financial ability means different things to different people at different stages of life. For the young, it often means acquiring the ability to afford things they want, such as cars, homes, and high-definition flat screen televisions. For the middle-aged, it means saving enough to provide for a comfortable retirement. And so on. One frustrated physician once vented to us that some of his patients seemed "to care more about their wallets than about their health." This is often true. Until poor health arrives, many people feel the need to achieve financial ability more intensely than they feel the need to be healthy.
4. Protect my assets from being taken or destroyed. This is the raison d'être of most types of insurance. We hire property and casualty insurance, mortgage insurance, and auto insurance to protect the value of important assets from being taken or destroyed in the event of theft, accident, or disaster.
5. Help me get paid fairly for products and services rendered. Getting paid at a profitable level is essential to the financial viability of hospitals, physicians, and networks. Providers of solution shop services need to receive fee-for-service payment; those providing value-adding process services can be paid for outcomes; and network facilitators need membership fees. All providers want to be paid in ways that account not just for costs, but for the value of the outcomes achieved—and to be paid promptly with minimal harassment. Similarly, makers of diagnostics, therapeutics, equipment, and other supplies need to remain sufficiently profitable to continue innovating new solutions.
6. Help me cost-effectively attract and retain the best possible employees, and make them as productive as possible. This job-to-be-done was the primary rationale for putting employers in the business of covering employees' health-care costs in the first place. This job persists today—and is experienced more urgently, in many companies, than in the past.
7. Help me avoid paying for unnecessary services.40 Health assistance companies, whether for-profit (e.g., Aetna, UnitedHealth), not-for-profit (BCBS), or government (Medicare and Medicaid in the U.S., national systems elsewhere), need to remain financially viable. A key lever for doing this is to pay only for medically appropriate services. This often pits insurers in an adversarial role against providers, whose job is also to make a profit and who submit claims at what they believe are fair prices. A key point of cost control for insurance companies is to deny, delay, or make partial payments—engaging in a costly tit-for-tat with providers that adds significant costs to health care.
8. Help me to win votes. One way politicians stay in office is by plausibly promising more, while still balancing the budget. But if balancing governmental budgets is hard now, we haven't seen anything yet. As we noted earlier, the trajectory of health-care cost inflation already has rendered nearly every local government technically bankrupt. And the cost of health coverage for the burgeoning population of the elderly will balloon America's Medicare costs to the point that 20 years hence there will be no room in the federal budget for anything except defense, Medicaid, and Medicare. Yet it seems that every significant attempt to reign in these exploding costs unleashes a maelstrom of electoral malice, intimidating even the most courageous politicians, thereby preventing this job from getting done.
In Figure 7.4 we've arrayed our assessment of how well the available alternative payment mechanisms do each of these jobs. We've again used Consumer Reports style of circles to rate the performance of each product (where a filled-in circle signifies that the product does that job excellently, and an empty circle depicts the product performing poorly). These ratings represent our synthesis of information gathered from hundreds of hours spent in interviews, industry meetings, and study of published reports.
Note that none of the systems does the first job excellently: "Help me to become healthy." The reason is that the payments mechanism doesn't heal. At best it constitutes no barrier to access. In our calculus, therefore, the best rating that can be given to a payments program for this job is a neutral one. Capitation in a nonintegrated system does this the worst, because it restricts access. We also rate national health plans as neutral. In many of these plans, entry-level care is more readily available than in the United States. But access to specialist care is often tightly rationed. Integrated capitation and the HDI-HSA combination score best on "help me to maintain my health" because they embody financial motivation to do this job.
FIGURE 7.4 Assessment of how well alternative systems get the job done
Fee-for-service, capitation, and nationalized health plans pay for health costs, but only the combination of HDI and HSAs has a mechanism to help consumers build financial ability. We rate nationalized health systems poorly for the "pay me for services rendered" job. The reason is that while collecting a paycheck from the national health service is simple, caregivers are typically paid at modest levels, compared to those in private practice.
The table in Figure 7.4 suggests that the three historically dominant forms of health coverage—fee-for-service (which includes Medicare and Medicaid), capitation in nonintegrated health systems, and nationalized health systems—each suffer from fatal flaws, in that they do one or more crucial jobs very poorly. FFS, in particular, rates badly on the jobs that politicians and insurance companies need to do, because they put no brake on the unbridled escalation of health-care costs. In contrast, the combination of Health Savings Accounts (HSAs) and HighDeductible Insurance (HDI) seems to offer a set of experiences and features that, while not perfect, do more jobs better than traditional alternatives.41 Capitation within integrated providers (integrated capitation) gets the best overall score.
If integrated capitation and the HSA-HDI combination rate so strongly, why has the switch to these new mechanisms been so sluggish? The answer for integrated capitation is obvious: capitation only works in an integrated provider system, and integrated providers currently account for about 5 percent of people covered in America.42 When capitation is attempted in a nonintegrated system, it places independent businesspeople in a zero-sum game, where one doctor's gain becomes another's loss. As the advantages of integrated providers become apparent among firms attempting to build a disruptive value network in health care, we would expect integrated capitation to grow in popularity.
There are two reasons why the uptake on the HDI-HSA cocktail has been slower than some reformers envisioned.
First, they have typically been offered in isolation—ignoring the mandate for creating a new disruptive value network that we discussed in Chapter 6, and without the business model innovations discussed in Chapters 3 through 5 that make week-to-week access to caregivers convenient and affordable. Comprehensive health plans have long shielded us from the real costs of health-care services. When the copayment is only $10, most of us have been willing to put up with the inefficiencies and poor service that the typical physician's practice offers. For $10, after all, what more could you expect? But when patients suddenly have to pay the doctor the full $150 cost of the visit out of their HSAs, they become very conscious of the value and convenience of what they're buying. And they rarely like what they see. For this reason, simultaneous innovations like retail clinics are crucial to satisfaction with HSAs. If we're going to ask people to start paying for their health care, we'd better make it affordable and convenient.
The second reason the rate of adoption has been more modest than anticipated is related to the first. One rule in launching disruptive innovations is that the initial customers should be non-consumers rather than users of the traditional products—because the only way the new will unseat the old among the original customer set is if the new performs better along the metrics of performance valued by those who use the old product.43 Many companies have simply offered HSAs as one option on a menu of health plans open to employees. In doing this, they pit HSAs head-on against sustaining competition from existing products—and history has shown that such strategies rarely succeed. When invited to compare HSAs versus comprehensive fee-for-service plans on a feature-by-feature basis, HSAs don't stack up well to the untrained eye. Employees compare the copays levied by their comprehensive plans, versus the risk that HSAs might entail an "out-of-pocket gap" of the sort depicted as the middle band in Figure 7.3. Many have consequently opted to "play it safe"—and stick with conventional comprehensive coverage.
Where paired HDI and HSAs have disruptively taken root is in small and start-up companies that can't afford the luxury of traditional comprehensive insurance. There is a direct analogy between this and the pattern by which 401(k) retirement investment plans disrupted traditional pension plans in the 1980s. Pension plans first appeared on the American benefits landscape in the early 1950s. For the first three decades, most workers' retirement benefits such as pensions and Social Security were structured as defined benefit plans. These plans defined the benefits retirees would receive upon retirement—a stream of payments after retirement that, in theory at least, would flow from cash being stockpiled during the employee's working life by the company or the government in a pension investment fund. Most small and start-up companies could not offer pension plans—their employees were left to save for retirement as best they could.
Beginning in the 1980s, Individual Retirement Accounts and 401(k) plans emerged. In these defined contribution plans, employers that had not been able to offer generous pension plans defined how much money they would contribute—which typically was much less than major companies were paying into their employee pension plans at the time. Because employees could contribute pretax money into these plans too, some employers offered to match employee contributions, and employees working in smaller companies could still save for retirement in a tax-advantaged way.
In contrast to defined benefit plans, in defined contribution schemes the employees are responsible for investing their retirement funds as they accumulate. The benefits that ultimately flow from the accounts depend on how frugally and well the employees contribute to and manage these accounts. When measured by the metrics of "quality" in the world of defined-benefit pension plans, 401(k) accounts don't stack up. But like all disruptions, these plans have proven to be enormously popular among nonconsumers of traditional pension benefits—because the 401(k) is infinitely better than nothing.44 Fewer and fewer companies now offer traditional pension plans as a result of this disruption.
Traditional comprehensive health plans—whether fee-for-service (including Medicare), capitation, or nationalized medicine—are defined benefit plans similar to pension plans. They focus on what employees and their families get out of the plan; and the assumption has been that employers and the government will put whatever is required into these plans in order to get the defined benefits out. Health savings accounts are defined contribution plans. The focus is on what employers and employees put into the plans and keep there. The benefits that participants are able to take out depend on how conscientiously they contribute, how carefully they manage what is put in, and how judiciously they use health-care services. We expect that, because of their disruptive character, these plans will take root first among companies and individuals that have struggled to pay the full cost of traditional comprehensive plans—small businesses, start-ups, and the self-employed—and then move disruptively up-market.
The perception today of many observers that the HDI-HSA combination will never gain enough traction to sweep the health-care world is caused by the S-curve pattern by which nearly all new technologies substitute for the old. When a new approach substitutes for an old one because it has a technological or economic advantage, the pattern of substitution almost always follows an S-curve,45 as depicted on the left side of Figure 7.5. The vertical axis measures the percent of the market for which the new approach accounts. The S-curves are sometimes steep, at other times gradual. But almost always disruptions follow this pattern: the initial substitution pace is slow, then it steepens dramatically, and finally asymptotically saturates 100 percent of the market.
A persistent problem emerges for the incumbent industry leaders when one of these substitutions occurs, however. When the nascent technology accounts for only a tiny fraction of the total market (they're on the flat part at the bottom of the S-curve), the incumbent leaders project linearly into the future and conclude that there is no need to worry about the new approach because it will not be important for a long time. But then the world flips suddenly, crippling the established companies. For example, after a decade of incubation on the curve's flat portion, digital photography flipped on the film companies very rapidly. The result? Polaroid is gone. Agfa is gone. Fuji is seriously struggling. Kodak alone caught the wave—and it's been a rough ride.46
You might think that companies would learn from this experience, but the S-curve pattern of adoption begs a vexing question: if I'm on the initial flat portion of the curve, how can I know whether my world will flip to the new approach next year or in 10 years? It turns out there is a way to forecast the flip. First, as shown on the right side of Figure 7.5, one must plot on the vertical axis the ratio of percentage market share held by the new, divided by the percentage share held by the old (if each has 50 percent share, this ratio will be 1.0). Second, the vertical axis needs to be arrayed on a logarithmic scale—so that .0001, .001, .01, 0.1, 1.0, and 10.0 are all equidistant, as can be seen in the graph. When plotted in this way, the data always falls on a straight line. Sometimes the line slopes upward steeply, and sometimes it is more gradual. But it is always straight. The reason is that the mathematics "linearizes" the S-curve.
FIGURE 7.5 Patterns by which the new substitutes for the old
As a result, you get a pretty good sense of the slope of the linearized curve, even when the new approach accounts for only 2 to 3 percent of the total. That makes it easy to extend the line into the future to obtain a general sense of when the new innovation will account for 25 percent, 50 percent, and 90 percent of the total. We call this line a "substitution curve."
Figure 7.6 shows the pace of substitution of HSAs and HDI, versus conventional private health plans.47 The new instruments now account for about 3.1 percent of the total number of individuals covered through private insurance.48 This suggests, however, that by 2010, HDI-HSAs will account for 10 percent of the market. The 50 percent mark will be hit in about 2013. They will asymptotically approach 90 percent by about 2016 or so. In other words, we've been on the flat part of the S-curve for several years, but we seem to be approaching the steep portion of the ramp quite quickly.
FIGURE 7.6 Past and future substitution of HSAs and HDI for conventional private health plans
Note that even after HDI-HSAs supplant today's comprehensive health coverage, solution shops will need to be paid on a fee-for-service basis; value-adding process providers can be compensated on a fee-for-outcome basis; and patient management and patient network providers can be paid on a fee-for-membership basis. They can adopt appropriate profit formulas, independent of the insurance-reimbursement method in use.
When different products compete for the same job, those products often converge—and as a consequence we expect that 401(k) and HSA plans will converge into two different pages on the account statements that we receive from financial management companies like Fidelity Investments.
By way of illustration, the BlackBerry, made by Research In Motion, Ltd., has been highly successful because it was designed to do a job that lots of people needed to do—to be productive in small snippets of time. The BlackBerry actually competes for this job against the cellular telephone in the mind of many customers—because placing a phone call is another way to be productive in a small snippet of time. We predicted in an article written in 2001 that as a result, the BlackBerry and the cellular telephone would merge into a single product—because at that time customers had to carry both products and choose between them whenever they had that job of "help me be productive in this small snippet of time."49 The BlackBerry subsequently has incorporated voice capabilities, and most cellular phones now offer wireless messaging, games, and personal organizer functions.
The principle is: when two different products are being hired for two different jobs that arise at different points in time and space in the lives of customers, we expect those products to remain independent of each other. Companies that offer both will not have an advantage over companies that focus on one or the other. But if the job(s) that different products are being hired to do arises at the same point in time and space in the life of a customer, then we expect companies that offer both products to build a competitive advantage. This is why, for example, companies that seek to become "financial supermarkets" by offering the full range of products customers could possible need—including checking accounts, savings accounts, credit cards, brokerage services, life insurance, consumer loans, and mortgage loans—typically fail. The jobs that each of these are hired to do arise at different points in a customer's life. In contrast, the reason why gasoline stations and convenience stores have converged is that the desire to fill up on junk food and the need to fill up with gasoline arise at the same time for many customers.
What this suggests is that companies like Fidelity Investments, whose 401(k) management services are positioned on the job of "help me achieve financial ability," as it was put in Figure 7.4, are likely to begin offering HSAs and market them to customers as fulfilling the same job as the 401(k). They will begin reporting their status in the same monthly statement to their customers. Since Fidelity's business model relies on a percentage fee of assets under management, they have an interest in helping customers maintain their health and keep long-term medical expenses low through appropriate preventive care. This type of integration between financial services and health care will accelerate as companies realize the interdependence of these products and continue to merge their features.
Indeed, in order to minimize administrative costs for HSAs, some HSA account holders are already given debit cards to use when purchasing health-care products and services—so that with no reimbursement paperwork required, owners simply swipe their HSA card, and special bar codes will verify the eligibility of the product or service to be purchased with HSA funds.50
Just as the FDIC, Federal Reserve Bank, and Securities and Exchange Commission oversee the operations of commercial and investment banks, a "Federal Health Assistance Commission" will need to set standards for financial health and customer service. Policies must be portable—so that as we change employers we can take our coverage and accounts with us. Different employers can choose to contribute different portions to the cost of our HDI and HSA accounts, but the accounts themselves must be personal and portable.
When politicians, advocacy groups, and emergency department administrators huddle to discuss the growing cost of health care, the increasing number of Americans who lack insurance coverage for catastrophic care is frequently fingered both as a culprit and a result. Some states, such as Massachusetts, have recently enacted rules requiring that everyone be covered by health insurance.51 Employers above a minimum size are required to purchase insurance for their employees. Smaller employers that cannot afford to purchase such insurance can receive government assistance in doing so; and low-income individuals can apply for assistance in purchasing their own insurance. To make these insurance products affordable, they typically have very high deductibles—creating the rather silly situation where the poor are forced to be "covered" by paying for mandated catastrophic insurance whose threshold is so high that the vast majority of them will be unable to use it.
The medical expenses foremost on the minds of most of the uninsured are not those resulting from catastrophic hospitalization. Rather, they are the costs of day-to-day care for acute infectious diseases and routine medications. When we mandate high-deductible insurance coverage and yet allow physicians' groups to block the licensing of affordable retail clinics in poor neighborhoods, those we are hoping to help have little access to the noncatastrophic care so essential to their daily well-being. Dr. Marcia Angell, former executive editor of the New England Journal of Medicine, calls this situation universal "coverage without care."52
There is a solution, however. Just as certain employers have created incentives for accumulating retirement savings by matching, dollar-for-dollar, employees' contributions to their 401(k) accounts, governments could do the same with HSAs—matching by formula contributions that low-income citizens make to their HSAs. In addition, by fostering low-cost disruptive business models such as retail clinics and patient networks that can be paired with payment mechanisms such as joint-contribution HSAs, governments can make a significant dent in the persistent problem called the "uninsured poor." The solution for the uninsured poor isn't just to help them afford health care. It must also make care affordable.
1. The Federal Register is a federal government publication that includes official transactions of the U.S. Congress, as well as all federal agencies.
2. We will return to the topic of regulations in health care in Chapter 11.
3. Two important recent works make the case for consumer value. See Herzlinger, Regina, Consumer-Driven Health Care: Implications for Providers, Payers, and Policy-Makers (San Francisco: Jossey-Bass, 2004); and Porter, Michael, and Elizabeth Teisberg, Redefining Health Care: Creating Value-Based Competition on Results (Boston, Massachusetts: Harvard Business School Press, 2006). Both thoroughly explore the issues of creating value that can be measured. We recommend their work to all serious students of these problems.
4. Medicare rates must also fit within the federal budget, and therefore often have little to do with the value delivered in return. Typically, commercial insurers will use Medicare rates as a baseline for setting their own payment schedules, thereby compounding the problem.
5. The motivation can even lead to questionable care, as evident by the investigation in 2002 of Redding Medical Center in California, where doctors were accused of performing unnecessary tests and surgeries. See Eichenwald, Kurt, "Operating Profits: Mining Medicare; How One Hospital Benefited from Questionable Surgery," New York Times, August 12, 2003; and Gaul, Gilbert M., "At California Hospital, Red Flags and an FBI Raid," Washington Post, July 25, 2005, A09.
6. Horwitz, Jill R., "Making Profits and Providing Care: Comparing Nonprofit, For-Profit, and Government Hospitals," Health Affairs, vol. 24, no. 3 (2005): 790–801.
7. Falk, Isadore S., et al., The Cost of Medical Care (Chicago: University of Chicago Press, 1933).
8. The first incarnation of Blue Cross was in 1929 at Baylor University, which offered a health plan to Dallas area teachers that guaranteed 21 days of hospital care for six dollars per year. The origin of Blue Shield was the California Physicians' Service, which in 1939 began offering physician services for $1.70 per month to low-wage employees. These prepaid plans were also a response to a contemporary push to adopt compulsory health insurance nationwide.
9. Thomasson, Melissa, "Health Insurance in the United States," EH. Net Encyclopedia, edited by Robert Whaples, April 18, 2003. Accessed from http://eh.net/encyclopedia/article/thomasson.insurance.health.us.
10. Stabilization Act of 1942. Title 50, Appendix—"War and National Defense," October 2, 1942.
11. The history of HMOs can be traced back much earlier, though Kaiser's predecessors were more limited in scope. In 1930, Dr. George M. Mackenzie of the Mary Imogene Bassett Hospital (now part of Bassett Health Care) implemented a program in which full medical coverage was provided in exchange for an annual premium of $25 per individual. See http://www.bassett.org/history.cfm for more information.
12. Johnson, Harry M., "Major Medical Expense Insurance," The Journal of Risk and Insurance, 32, no. 2 (1965), 211–36.
13. See Follmann, Joseph F., "The Growth of Group Health Insurance," Journal of Risk and Insurance, vol. 32, no. 1 (March 1965), 105–12; Johnson, Harry M., "Major Medical Expense Insurance," Journal of Risk and Insurance, vol. 32, no. 2 (June 1965), 211–36; Hallman, G. Victor, "True Catastrophe Medical Expense Insurance," Journal of Risk and Insurance, vol. 39, no. 1 (March 1972), 1–16; and Wilensky, Gail R., et al., "Variations in Health Insurance Coverage: Benefits vs. Premiums," Milbank Memorial Fund Quarterly. Health and Society, vol. 62, no. 1 (Winter 1984), 53–81. More information on the HMO Act can be found at http://www.hpolicy.duke.edu/cyberexchange/Regulate/CHSR/PDFs/I-1-HMO%20Act%20of%201973.pdf.
14. John Wennberg, Elliott Fisher, and their colleagues at Dartmouth University pioneered this research on the drivers of health-care spending. See also Pearlstein, Steven, "A Better Way to Spread the Health—and the Wealth," Washington Post, February 8, 2006, D01.
15. This opposition from the AMA led to the compromise in which physician services were carved out of medical coverage that would be provided by Medicare. Therefore, Medicare would only cover hospital services, funded by payroll deductions, made available to all Social Security beneficiaries, and which later came to be known as Part A of Medicare. Eventually, physician and outpatient services were incorporated through new legislation that created Part B, which was funded by monthly premiums and had optional enrollment.
16. Our readers will note that we loosely use the term "profits" instead of the term "surplus" that is more commonly associated with "nonprofit" organizations. We deliberately do this out of a belief that the categorization scheme of for-profit vs. nonprofit is relevant only for tax purposes. From a managerial point of view, executives of both types of organizations strive to generate sufficient surplus revenue beyond expenses so they can sustain the health of their enterprise as they try to satisfy the needs of their multiple stakeholders.
17. "A Conversation with Matthew Holt," Managed Care, July 2004.
18. It is interesting to note that Kaiser has had poor results historically when trying to expand its business model elsewhere, even though it remains popular in California. In part, this may be due to Kaiser's difficulty in creating a truly integrated capitation system in communities where providers have largely been independent businesspeople.
19. Though many integrated organizations still tied physician salaries (typically bonuses) to referrals, for example, most of the risk in managing capitation fell upon the organization rather than the independent provider.
20. We thank Dr. John Hsieh for sharing with us his insights and clinical expertise in nephrology.
21. "The History of Dialysis," DaVita Web site. Accessed from http://www.davita.com/dialysis/motivational/a/197.
22. JPMorgan North America Equity Research, October 23, 2006. Much of the background history and data for this case comes from JPMorgan's NxStage analyst report.
23. Blagg, Christopher, "Having Options: Home Hemodialysis," RENALIFE 18, no. 5 (2003); and Blagg, Christopher, "A Brief History of Home Hemodialysis," Adv Ren Replace Ther, 1996 Apr:3(2):99–105.
24. NxStage Web site, accessed from http://www.nxstage.com/acute_renal_care/Products/index.cfm.
25. Lee, Helen, "Cost Analysis of Ongoing Care of Patients with End-Stage Renal Disease: The Impact of Dialysis Modality and Dialysis Access," Am J Kidney Dis 2002;40(3):611–22; and McFarlane, P. A., et al., "The Quality of Life and Cost Utility of Home Nocturnal and Conventional In-Center Hemodialysis," Kidney International, Sep. 2003, 64(3):1004–11.
26. This estimate is based on the U.S. Renal Data System (USRDS) 2007 Annual Data Report. Accessed from http://www.usrds.org/adr.htm. Calculation was made by combining the total Medicare costs for outpatient hemodialysis and for dialysis capitation paid to physicians and suppliers in 2005 (excluding inpatient dialysis). Actual savings would be lower due to some patients being unable to manage home hemodialysis, while savings would be higher after including cost savings of non-Medicare programs as well as possible improvement in clinical outcomes.
27. Based on data from USRDS 2007 Annual Data Report, comparing the total outpatient hemodialysis costs paid by Medicare in 1991 versus 2005.
28. USRDS 2006 Annual Data Report.
29. JPMorgan North America Equity Research, op. cit.
30. See Roake, J., "Withholding and Withdrawing Therapy: Humanity, Human Rights and Access to Renal Dialysis," New Zealand Medical Journal, June 6, 2003, vol. 116, no. 1175; and Busko, M., "Home Hemodialysis Prevalence Varies Greatly by County, Could Be Much Higher," Medscape Medical News, August 14, 2006. Accessed from http://www.medscape.com/viewarticle/542738.
31. Mayo Clinic National Symposium on Health Care Reform. Rochester, Minnesota, May 21–23, 2006.
32. In 1989, Medicare adopted a payment schedule based on the Resource-Based Relative Value Scale, a system that hoped to give rationality to payments based on value delivered. Created at Harvard University by William Hsiao and a multidisciplinary group of researchers, the RVS payment schedule determined appropriate reimbursement levels based on the relative value of each procedure and service. However, the RVS system was heavily politicized and distorted, and ultimately unsuccessful in tying reimbursement to actual value.
33. Median operating margins (which exclude investment income) of hospitals were 2 percent in 2004, according to Appleby, Julie, "Hospitals' Profit Margin Hits 6-Year High in 2004," USA Today, Jan. 4, 2006, 3B. According to the American Hospital Association, profit margins at hospitals recently moved from negative 16 to 2 percent. Medicare Payment Advisory Commission Public Meeting, Washington, DC: MEDPac, March 11, 2005, accessed from http://www.aha.org/aha/content/2005/pdf/050311medpactranscript.pdf on August 31, 2008.
34. This decision was certainly not malicious, however, and in fact was rooted in good intentions. Insurers reasoned that encouraging their customers to see their physicians earlier for low-cost, routine visits could possibly prevent higher-cost, downstream events like hospitalization. Yet, in retrospect, this was also the start to moving patients further from having to make any real trade-offs in their health purchasing decisions, as "first-dollar coverage" increasingly became the norm.
35. Some employers will find the cost of a high-deductible plan to be even less, in which case they would have the option of keeping some of the cost savings or returning some of it to employees in the form of higher wages or other benefits.
36. The maximum contribution for family coverage in 2008 was $5,800. This is adjusted yearly to keep pace with inflation, and employers may choose to increase their contribution levels accordingly. The HSA is the only employee benefit account that has a "triple tax advantage"—money deposited into HSAs is prestate, -federal, and -FICA tax, grows tax-free, and can be spent tax-free on health-care-related expenses.
37. Although the HSA-HDI combination in this scenario may not be advantageous for the chronically ill and for families with high medical expenses if their yearly health-care costs already exceed $5,000 annually and the HDI carries a deductible of greater than $5,000. However, if the deductible is less than $5,000, they would also be better off under this plan. The minimum deductible to qualify as HDI in 2007 was $2,200.
38. Year-to-year fluctuations in growth rate can occur of course. However, savings vehicles like health savings accounts and retirement accounts rely on a much longer time horizon, as well as appropriate diversification. We encourage readers to base their decisions upon the historical compound annual growth rate of well-diversified, market-based investments, in which case our estimate of 8 to 10 percent growth will hold.
39. We thank Mr. John Kaegi, chief marketing officer of Blue Cross and Blue Shield of Florida, for sharing with us his very capable body of research whose aim has been to identify what these jobs-to-be-done are.
40. This includes medically inappropriate services that are supply-sensitive (as the Dartmouth Institute for Health Policy & Clinical Practice would describe it), so-called "never events," and fraud.
41. Note that HSAs fulfill some of these jobs better than its contemporary insurance vehicle, the Health Reimbursement Account. HRAs are not portable, meaning that funds remain with the employers when employees change jobs, and thus they fail to help consumers achieve financial stability by remaining healthy. It also does not always fulfill the jobs of (1) cost effectiveness for employers and (2) avoiding paying for unnecessary services for insurers. Some employers choose not to rollover unused HRA funds from year to year, creating a perverse incentive for employees to use up all their remaining HRA dollars before they go to "waste" at the end of each year. The same incentive exists when employees plan to leave the company.
42. State Regulatory Experience with Provider-Sponsored Organizations, Falls Church: The Lewin Group, Inc., June 27, 1997.
43. The findings on this are summarized in Chapter 3 of Christensen, Clayton, The Innovator's Dilemma (Boston, Massachusetts: Harvard Business School Press, 1997); and Chapter 4 of Christensen, Clayton, and Michael E. Raynor, The Innovator's Solution (Harvard Business School Press, 2003). In addition, a book co-written by our colleagues at Innosight LLC discusses how to spot nonconsumption and what to do about it; see Anthony, Scott D., et al, The Innovator's Guide to Growth (Harvard Business School Press, 2008).
44. One other advantage is that 401(k)s and HSAs are both portable. Under traditional pension plans, employees were typically tied to an employer for five to 10 years before their pension funds were vested.
45. Everett Rogers, Diffusion of Innovations (New York: Free Press, 1962). Richard Foster of McKinsey has also studied this phenomenon thoroughly. See The Attacker's Advantage (New York: Summit Books, 1986).
46. Another example is AT&T, which relied on a McKinsey study that in 1984 advised them there would be fewer than 1 million wireless phone units by 2000. There were 740 million—reminiscent of IBM CEO Thomas Watson's forecast that the world would not need very many computers.
47. Data for HSA enrollment are from the AHIP (America's Health Insurance Plans) Center for Policy and Research (http://www.ahipresearch.org/). There were 438,000 enrollees in HSA plans in September 2004; 1,031,000 in March 2005; 3,168,000 in January 2006; 4,532,000 in January 2007; and 6,118,000 in January 2008. These figures do not include individuals with HDI but no HSA. Data for Private Health Insurance coverage are from the U.S. Census Bureau Current Population Survey Annual Social and Economic Supplement (http://www.census.gov/hhes/www/cpstc/cps_table_creator.html). There were 199,870,585 lives covered by private insurance in 2004; 200,923,910 in 2005; 201,167,391 in 2006; 201,690,112 in 2007; and 201,990,660 in 2008.
48. Some individuals are simultaneously covered by other insurers, such as Medicare and the Veterans Health Administration.
49. This was incorporated into Chapter 3 of Christensen, Clayton M., and Michael E. Raynor, The Innovator's Solution: Creating and Sustaining Successful Growth (Boston, Massachusetts: Harvard Business School Press, 2003).
50. Rubenstein, Sarah, "Patients Become Consumers," Wall Street Journal, December 28, 2005.
51. "Health Care Access and Affordability Conference Committee Report," Commonwealth of Massachusetts, April 3, 2006. Accessed from http://www.mass.gov/legis/summary.pdf.
52. Kuttner, Robert, "A Health Law with Holes," Boston.com, January 28, 2008. Boston.com is an online product of the Boston Globe.
