Sick To Death > Chapter 4 > Barriers to Reform
Much of the resistance to reform arises from the "usual suspects" of self-interest and inertia among those who make a living under the current arrangements, as well as inattention engendered by a widespread distaste for reminders of implacable mortality. A few issues entail conflicts among deeply held political and ethical values; society's unwillingness to deal with these both obstruct more productive resolutions and set intractable challenges for reform efforts. The three issues that I highlight here are these:
Effective new drugs and devices promise to overwhelm the health-care budget. As just one example among many, the increasing use of implanted cardioverter defibrillators (called ICDs) will generate remarkable expenses within just a few years. Under the restricted criteria from the Multicenter Automatic Defibrillator Implantation Trial II, three or four million Americans already qualify for an implanted defibrillator that can stop most life-ending heart irregularities (Moss et al. 2002). Another six hundred thousand qualify each year. One estimate put the costs per year of life extended at $125,000 (Greene 2000). Each device costs $25,000, and implanting and regulating it cost about that much every year. Even if only the new cases actually got the device, and if life span with the device averages a few years, the costs would be at least $60 billion per year (600,000 people/year × $100,000/person = $60 billion/year). Most of these costs are in Medicare, since the heart condition mostly affects older people. Even with these deliberately conservative estimates, defibrillators would account for one-eighth of the current Medicare budget (Center for Medicare and Medicaid Services 2002) - for a device that was not included in budget projections just a year or two ago. ICDs are effective and, on average, extend life. They also make it much less likely that dying will be sudden. For some patients, the other ways they could die may well be rather more difficult.
This kind of high-cost but effective treatment challenges the social fabric in new ways. Medicare has long been mandated to provide all advantageous treatments that are proven effective for particular conditions. No part of the approval, pricing, or coverage process even allows anyone to make decisions to restrict availability on the basis of costs once the treatment is shown to be more effective than older alternatives. New treatments like defibrillators may finally have brought society to the breaking point. In the 1980s, the nation absorbed the expenses of dialysis and kidney transplants, despite unexpectedly high costs. In comparison with those who could use implantable defibrillators and genetically engineered drugs, the potential pool of renal failure patients was small, and patients were often young and otherwise healthy. The pool of people eligible for the newer developments is much larger, older, and sicker.
While implantable defibrillators are a particularly visible and costly example, many other treatments raise the same questions. Lung volume reduction surgery (http://cms.hhs.gov/ncdr/memo.asp?id = 96), left ventricular assist devices (http://cms.hhs.gov/ncdr/memo.asp?id = 79), and genetically engineered drugs are new entries in the high-cost sweepstakes; but chemotherapy for advanced cancer (Emanuel et al. 2003), intensive-care-unit support for the very old and frail, and expanding use of erythropoetin to stimulate blood production are existing examples of the same widespread use with small gain. Left ventricular assist devices, on average, extend life less than a year at a cost of over $2,500,000 (Gillick 2004). Perhaps even the affected patients might have higher priorities for that sizable fund (Smith et al. 2003).
Talk of "rationing" health care has an ominous ring to the public and, as a result, to politicians. Nevertheless, more treatments now successfully extend life for people who have verylimited life spans and life possibilities. All of us, and our political leadership especially, must address questions of balance and merit (Callahan 1998). Failing to do so forces us to go on as we are now - unable to talk about acceptable deaths, extended but undesirable life conditions, or priorities among social investments.
If we cannot limit the distortions possible from providing very high-cost treatments mostly aimed at people in the last tenth of their lives, then we may shave away or deny funding for aides, nurses, housing, and other fundamentals. Denying a particular medication or treatment that patients or families want often proves to be contentious and public. Reducing those services that are hard to track or to demand, or whose benefit is not so dramatically apparent, is much easier. For example, hospices that once provided four to six hours per day of home health aide coverage have gradually cut back to the occasional one- or two-hour session. No one notices this, but families would be up in arms if the hospice instead restricted use of high-cost pain medications. So the challenge now is to establish a method by which the federal budget for care of those with fatal chronic illnesses can match the dominant problems and concerns of patients and families, even if this framework means that very costly treatments are sometimes not readily available to some patients who might benefit.
Accomplishing this aim will be quite difficult in our culture. Americans are proud of making every proven medical treatment available to all, at least in Medicare. Medicare cannot lawfully refuse to cover a treatment because it is costly. The urgent question that these costly treatments set before us is whether the merit for patients is worth the cost to society.
As we try to learn how to generate an answer, what approaches might work better? First, we could tailor services to evidence indicating what the covered population values, allowing for some latitude for patients with unusual preferences. For example, in 2003, geriatricians debated whether residents in nursing homes with modestly elevated lipid profiles should have lipid-lowering drugs, given that new data showed slightly improved survival, even in old age. In the usual course of things, that argument would simply carry the day, and a standard would be set to use lipid-lowering drugs even in advanced old age and even for persons with severe cognitive failure. But such a decision should at least weigh the preferences of the residents, their families, and their hands-on caregivers. At prices in 2003, the costs for a two- hundred-bed nursing home would be about $150,000 per year. Suppose we asked a broad set of those most concerned - the residents, families, and nursing assistants - to list the uses of $150,000 that would enhance the lives of the residents. As it turns out, no one with whom I have discussed this situation believes that affected parties would ever list using the money on lipid-lowering drugs. My point here is that data about the preferences of the parties most affected actually have almost no force in setting the care system's priorities. If we set out to make them salient, we could require evidence of these preferences in coverage decisions and standard setting. While the preferences of individual patients and families should be weighed in individual treatment decisions, I am most interested here in giving the aggregate preferences of affected groups of patients and families a role in shaping the care delivery system. Because not all options are feasible, we would serve patients well if the options that mattered most were also the most readily available.
A second approach would be to find ways to expect or require consideration of coexisting medical problems and shortened life expectancy in decisions about providing treatments. When Medicare considers coverage of implantable defibrillators, as described earlier, the data that policy makers use have evaluated the usefulness of the device in stopping death in people afflicted mostly with heart failure. The study patients must be in good enough health that they are not expected to die of something else in the near future. Yet actual patients ordinarily have other serious illnesses that limit life expectancy. Indeed, patients receiving ICDs are not generally warned that they will want to stop the defibrillator at some point before dying, either to avert the troubling firing of electric shocks when near death or to accept dying of heart rhythm irregularities rather than waiting for another, more difficult, course to death.
The nature of life with dementia, and the latitude allowed to family choices and advance planning that decide whether or not to treat to extend life, strikes many people as a troubling situation. We want to honor the dignity and worth of all human beings, but we also often find it unacceptable or at least unnecessary to make life longer once dementia has robbed people of their character, memories, and capabilities. While individuals and families struggle with these issues as they play out in a family member, we also need a workable social consensus, and approval of at least an accepted range of options. With half of people who die after age eighty-five having a cognitive deficit (Evans et al. 1989), society will have to make peace with a range of choices about treatments to prolong life with serious and progressive dementia. Indeed, it may be that assessment of people's perceptions of what to value in severe dementia would mostly find priorities of comfort, dignity, cleanliness, skin care, and consideration for a family's burdens - and not prolonged life.
The sentiments embodied in the Americans with Disabilities Act and other public policies to defend the interests of people with disabilities make it difficult for anyone except the affected person to weigh the value of extended life (National Council on Disability 2003). If a person with a particular medical problem and no disabilities would normally receive a given treatment, then that treatment must be made available to anyone with disabilities who also has that medical problem. As a protection against discrimination in health care, the Americans with Disabilities Act has been important. Its commitments also lead to use of costly and weighty treatments in elderly persons with serious and progressive disabilities like dementia. Demented nursing-home patients sometimes have implanted defibrillators because they have no family or advance directive that could instruct providers to forego the treatment. The dementia patient who goes through defibrillator implantation in order to put off a peaceable natural death ordinarily gets to live through a more prolonged decline and, overall, a much more difficult dying. In my view, this has to be among the more incomprehensible practices that well- meaning policies have engendered. Yet it is not easy to see how to disassemble this legal structure without harming the fragile endeavor to ensure the rights and access to health care for people with disabilities (Gostin 2003).
Thus, the priorities in this arena might include
Building a new and better care system tailored to long-term, eventually fatal chronic illness requires learning how to assess its strengths and weaknesses and how to arrange, finance, and regulate its services. That learning requires innovation, evaluation, and research. Some of this work is small-scale innovation that would usually be within the scope of approval by those managing programs or providing professional services: for example, trying out a new staffing pattern for nurses or a better way to ensure pain assessment and relief. Some of it involves broad social innovation, like changes in eligibility, coverage, or payment rates. Other parts of the work take the form of organized formal research, with control groups and defined interventions.
Society can take various actions to encourage or to discourage innovation. In the early years of the United States, public policies allowing people to leave bankruptcy behind and encouraging them to take risks on new investments helped engineer an era of invention and development. At present, substantial forces act to discourage innovation and encourage continuation of "accepted practices." Conventional liability litigation, for example, has been a strong force against change in health care generally, since there is often a period of heightened risk as a new approach takes root.
Increasingly, concern for the protection of human subjects complicates innovation as well. In order to ensure that human subjects of experimentation are knowledgeable about their risks (of harms and of loss of privacy) and that they consent freely, a set of rules and protections is in place, requiring substantial review of the research plan, the consent provisions, and the implementation of the research. The processes that govern research are weighty and often very slow (Lynn 2004; Casarett, Karlawish, and Sugarman 2000; Dubler and Bellin 2001).
Whether innovation, program redefinition, continuous quality improvement, and other reform endeavors count as research, and whether they should, is a subject of current concern (Lynn 2004; Lo and Groman 2003). If data-guided quality improvement must be regulated as research, the pace of change will be much slower than otherwise. If not, then society may need to articulate alternative arrangements in order to ensure that reformers treat patients equitably and with respect. In the absence of a clear process, the fears of liabilities and the demand to act conservatively could be major sources of delay and inaction.
Approaches to these problems might include