Demography, technology and choice

Although ‘popular myth’ (Reinhardt, Reference Reinhardt2003: 36) assimilates end-of-life issues to the doom-ridden demographic destiny that the growth of the aged population allegedly inflicts on health care systems and budgets, debunkers show that the costs and ‘character’ of end-of-life care derive not from when one dies but from how and where, which are, at least in principle, matters of individual preference, social choice and institutional design (Zweifel et al., Reference Zweifel, Felder and Meiers1999; White, Reference White2004; Gusmano and Allin, Reference Gusmano and Allin2011). Instead of tackling these challenges, however, US policymakers arguably opt for dodging and denial. As Daniel Callahan and others have observed, the US gives unqualified obedience to the imperatives of medical technology, the ‘beloved beast’ (Callahan, Reference Callahan2009) that protracts human suffering, wastes social resources, and prevents all concerned – patients, families, providers and policymakers – from knowing when to say when. On this view, the nation needs above all to get real – be more tough minded, affirm the need for hard, nay tragic, choices and start designing socially acceptable ways to make them.

A spectator of its health reform deliberations in 2009–2010 might conclude, however, that the nation is moving in precisely the opposite direction to what critics of the technological imperative recommend. The most conspicuous evidence for this conjecture is of course the reception evoked by a congressman's modest proposal that Medicare should start paying physicians for the time they spend counseling aged patients on their end-of-life options. Sarah Palin (former governor of Alaska and Republican John McCain's vice presidential nominee in his losing bid for the presidency in 2008), joined by a parade of other conservative ‘thought leaders’, attacked the initiative as a formula for ‘Death Panels’ (Blumenauer, Reference Blumenauer2009), and although the allegation won wide and deserved derision (and the “Lie of the Year” award for 2009 of the website Politifact) it threatened to derail the health reform project of the Obama administration in the waning months of 2009. In late December 2010, health reform legislation well in hand, the administration tried again, issuing rules under the authority of the Centers for Medicare and Medicaid Services that permitted Medicare payments to physicians who advised patients on end-of-life care (Pear, Reference Pear2010), whereupon the tempest resumed and Obama quickly reversed his decision.

This article explores noteworthy and rather more nuanced US developments between the poles of resolving to pull plugs and crusading to disempanel death. Contemplating American experience within the three policy ‘realms’ examined here – advance directives, hospice and palliative care and assisted suicide – suggests that beneath the often inflamed rhetoric and beyond the nation's apparent distaste for coherent collective deliberation on end-of-life practices and policies, the US is ‘stealing on insensibly’ (the phrase comes from John Donne's Devotions) toward end-of-life treatment options that respect and secure the preferences and decisions of patients and their loved ones.

Advance directives

Public policy ensures that in principle no American need be at the mercy of unwanted life-extending medical intervention. This principle has gained strength in both public and judicial opinion for almost 40 years as ‘landmark’ decisions – Quinlan (1976), Cruzan (1990) and Schiavo (2005) to cite the most prominent – have appeared roughly every decade and a half. The central issue now, note Murray and Jennings (Reference Murray and Jennings2005: S52), is less empowering individuals and their agents to make decisions than ‘how to enable them to make good decisions’.

The most familiar type of advance directive, the living will, was first enacted in California in 1976 and by 1992 was on the statute books of all 50 states. Since 1990, federal legislation (the Patient Self-Determination Act, which was included in PL 101-508, the Omnibus Budget Reconciliation Act of that year) has obliged medical providers to inform Medicare patients about their rights to make such documents. In 2010, President Barack Obama urged all Americans to execute a living will (and incurred mild reproof from cognoscenti for not speaking more broadly of ‘advance directives’).

Anyone with a computer can download a living will document – for free from Caring Connections, for $5 from Aging with Dignity and for a lifetime fee of $125 an organization called US Living Will Registry will supply the form and register the completed document online. One fills it out and signs the will, has it witnessed and notarized, and presumably then makes it available to family members and medical providers. The process is not entirely cut and dried: the bearer must remember where it is stored and to whom it is accessible, update it if one has a change of mind, and be sure to comply with the specific requirements of the state in which care is expected to take place (a possible complication for retirees with more than one home), but these caveats would seem to be manageable for much of the adult population. Bearers of living wills can spell out in as much detail as they please precisely what they want done or not done to them medically should they become incompetent to speak for themselves.

A range of arresting figures indicates both the demand for ‘voice’ about end-of-life care and the disconnect between demand and supply. A recent study in Maryland found that although ‘only 34% of respondents had an AD, 61% indicated that they have preferences about medical care in the event they are unable to make such decisions’, and of these ‘83% said it was very important that their preferences be carried out’ (Pollack et al., Reference Pollack, Morhaim and Williams2010: 59). Research by the Pew Research Center for the People and the Press found that in 2006, 69% of married people had talked with their spouse about their preferences for end-of-life care (up from 50% in 1990), nearly $$)(--><$>{3 \mathord{\left/ {\vphantom {3 4}} \right. \kern-\nulldelimiterspace} 4}<$><!--$$ of the public believes it is sometimes right to let a patient die and only 22% contends that doctors should always try to save a life, so one may wonder why only about 30% of the public has adopted a living will (Moreover, getting a living will to do its job is no simple matter: for example, 62% of those who have one fail to give it to their doctor, and 64% of dying patients’ living wills do not cover the clinical realities they face.) (Amednews.Com, 2009).

Researchers in Maryland report that the main reasons why people do not have advance directives are ‘unfamiliarity with them and the belief that it takes too much time, effort, or money to complete them’ (Pollack et al., Reference Pollack, Morhaim and Williams2010: 61). For those who are acquainted with the option and inclined to proceed the disconnect seems to center on two problematic features of the specificity a workable living will requires. A blogger warns that advanced directives are never a mere ‘check box’, but rather are ‘wickedly time consuming, nuanced and difficult’ (Health Beat Blog, 2009: 8). The University of Michigan Health System (2005) warns that “If you want your patient advocate to be able to refuse any treatment and let you die, you must say so specifically in the durable power document” (www.med.umich.edu). Given the complexity of organ systems, life circumstances and ambiguous future interplay between the two, people have trouble gazing into the future with enough perspicacity to envision the range of possible condition/treatment pairs they might want to rule out. Predictive abilities aside, moreover, the injunction to record one's preferences in highly specific detail doubtless aggravates the reluctance of no small part of the population to ponder and discuss end-of-life issues and therefore, ironically, may itself deter people from inquiring into and seeking to execute a living will.

The American Bar Association's Commission on Legal Problems of the Elderly has flatly branded as a ‘myth’ the notion that everyone ought to have a living will. The better type of advance directive, says the association, is durable power of attorney (POA). In this case, one completes legal documents that designate someone else – usually a relative or close friend – as one's ‘surrogate’ decision maker for medical matters, thus empowering him or her to exercise ‘substituted judgment’ when questions as to one's intent arise (POA documents can be coupled with living wills and the mechanics of completing the two are much the same).

POA arrangements do not entirely eliminate the problem of precise prediction, however. For example, if the signatory wants a surrogate to be able to stop treatment and let him or her die, this must be stated specifically in the POA document. And of course POAs presuppose a degree of trust in others in matters of life and death that not everyone is prepared to muster. They may also gain value from esoteric ‘insider’ savvy. For instance, in creating his medical POA documents, David, “a blogging RN with long experience in and around ICUs”, added to the boilerplate “significant verbiage on my own beliefs and philosophy, getting into issues that the average patient would have no clue about – the use of opiates to treat shortness of breath even at the cost of shortening life”, for example, and then “had a pretty good discussion with both my primary proxy – my wife – and my secondary proxy – a nurse friend”. Having done his best to ensure that his wishes would be followed, he remained “acutely aware that very few people in the world have the knowledge to do what I did” (posted in Health Beat Blog, 2009).

As David RN's case suggests, considerations broadly related to culture and social class may come into play in POAs and living wills alike. To benefit from them, one must be able to read and write, know that advanced directive options exist, have access to a computer, assemble witnesses, find and pay a notary, keep the advanced directive on hand and on record and present them to medical providers in a timely fashion. People who fail to meet these (admittedly modest) tests for reasons of poverty, social isolation, advancing dementia or otherwise may find themselves given ‘30 seconds in a busy lobby to read and sign a complex document’ (Pallimed, 2007) or assigned a court-appointed guardian. In the Bronx, notes Nancy Dubler (Reference Dubler2005: S22), providers find that patients are ‘almost uniformly uninterested’ in ADs. These patients, “many of them persons of color are interested in access to care, not in limiting care”. Deborah Carr (Reference Carr2012: A20) reports that people in the lowest quartile of assets are “only half as likely as those at the top of the assets ladder to have a living will, to appoint a health care proxy or to discuss their treatment options with loved ones”. Only about 15% of Americans have arranged medical POAs – half the (modest) number with living wills.

Advance directives aim to express an individual's demand (or lack thereof) for medical care at the end of life or in other emergencies. Whether providers opt to supply care that conforms to these formally revealed preferences is another matter, on which authorities disagree. One expert declares that ‘literally hundreds’ of studies show the inadequacies of ADs (C. Caleb Alexander quoted in Amednews.Com, 2009) but a recent paper in the New England Journal of Medicine paints a more positive picture (Silveira et al., Reference Silveira, Kim and Langa2010). Physicians caught between the patient's ‘autonomy’ and his or her ‘best interests’ may hesitate how to proceed.

The all-American fear of litigation may come into play: physicians are more afraid of legal repercussions from their failure to extend a life than from failure to end one (Health Beat Blog, 2009). And no American tale could be complete without at least one blatant example of meddling by religious ‘authorities’ – as for instance in the case in Tulsa, Oklahoma, in which a patient with a living will that ruled out artificial hydration or nutrition was given a feeding tube because her nephew, acting on a Catholic directive that changed the ‘presumption’ in favor of feeding tubes to an ‘obligation’ to provide them, refused to let the (Catholic) hospital withdraw treatment. (Catholic institutions in his district, decreed Bishop Edward J. Slattery, “will not honor an advance directive that is contrary to Catholic teaching.”) (Article in Tulsa World reported by Meyer, Reference Meyer2010.)

Even were litigation and religion purged from the picture, advance directives would fail to speak for themselves as definitively as their bearers might wish. Vague instructions in ADs “result in conversations that produce equally vague expressions of wishes such as ‘Do not keep me alive with machines’ or ‘Let me die if I am a vegetable’ ” (Hickman et al., Reference Hickman2005: S27). Bioethicist John Lantos observes that ‘current clinical circumstances’ may be subject to varying interpretations and of course may change as a patient wavers between life and death. Surrogates asked to define what the patient ‘would’ want done in these possibly ambiguous conditions, remarks medical professor Angel Fagerlin, face ‘undue pressure’ and may feel unsure what to do (Amednews.Com, 2009). Because physicians may not honor an advance directive, the American Bar Association cautions their bearers to talk it over with caregivers ‘before a crisis occurs’ (Carpenter, Reference Carpenter2010) – a bit of prudent advice that begs the question but elicits rare agreement between the lawyers and organized medicine. According to the American Medical Association, end-of-life decisions are best viewed as exercises in negotiation and consensus building. Thus:

• • Identify points of agreement about broad goals, as well as specific therapies

• • Negotiate points of conflict by exploring values

• • Defer to future discussions when necessary

• • Try time-limited trials of various approaches

• • Agree to disagree (Amednews.Com, January 5, 2009).

The notion that death is no less political than is life itself may bring macabre cheer to a political scientist, but those who faithfully download, notarize, file and the rest in hopes of ending or preempting conversation on the end-of-life care they will receive may be less than thrilled to learn that in fact they have merely set the stage for new conversations by others about their eventual medical fate.

As might be expected, the textual technology of end-of-life directives steadily evolves. Perhaps the most promising variation to date is the Physician Orders for Life-Sustaining Treatment (POLST). Instead of relying on the judgment of surrogates, the POLST is ‘an official medical order’ that ‘spells out specific treatment instructions and remains part of a patient's medical record’. On the statute books or under development in 33 states, POLST forms, which are signed by physicians, and usually present a choice among comfort care, limited intervention and full treatment, should be offered to people “when their physician would not be surprised if they died in the next year”, says Dr. Susan Tolle. Because paper-based forms may be misplaced, lost or overlooked, some states are now building electronic databases for POLSTs (Murray, Reference Murray2010: A5; Landro, Reference Landro2011: D1–D2).

Palliative and hospice care

The conversations that advanced directives trigger often center on the timing and nature of palliative care and hospice services. Hospices are programs that deploy interdisciplinary teams (nurses, social workers, home health workers and more) to deliver ‘comfort care’ (including palliation), but not curative medicine, usually (though not necessarily) in the home of a patient who is at or near the end of life. Palliative care is an umbrella term for services that aim to manage and reduce pain in patients who may concurrently get curative care and who may be, but need not be, terminally ill. Hospice care invariably entails palliation, but this latter transpires in settings that include but extend beyond hospices, notably hospitals.

As generic venues for rest and shelter, hospices date back to the crusades. They became ‘institutionalized’ in the 19th century in settings such as the Our Lady's Hospice in Dublin, created by the Religious Sisters of Charity in 1879 and began to expand internationally under the leadership of (among others) Dame Cicely Saunders, who founded a hospice (St Christopher's) in a London suburb. In the late 1960s, as Elizabeth Kubler-Ross's book, On Death and Dying, drew popular attention to ‘death with dignity’, Florence Wald, dean of Yale's School of Nursing, worked with Saunders to promote hospices in the United States. Important indicators steadily signaled that hospices were becoming ‘mainstream’ – for example, their inclusion as a guaranteed benefit in President Bill Clinton's failed reform plan of 1993, the issuance of a Hospice Care commemorative stamp in 1999, a national conference on access to hospice and palliative care in 2005, and, not least important, a growing body of research purporting to show that hospice care reduces the costs of care at the end of life (National Hospice and Palliative Care Organization, n.d.).

In the United States, hospice care has been authorized since 1983 to receive payments from Medicare (which is now the source of about $$)(--><$>{3 \mathord{\left/ {\vphantom {3 4}} \right. \kern-\nulldelimiterspace} 4}<$><!--$$ of funds for the sector) and state Medicaid programs also include hospice care of varying scope, as do most private health insurance plans. Hospice is a $10 billion industry that enrolled 1.4 million terminally ill Americans in 2007 (Meier and Beresford, Reference Meier and Beresford2009: 594) and is fortified by interest groups such as the National Hospice and Palliative Care Organization (created in 1978) and the American Academy of Hospice and Palliative Medicine (1988), which labor to protect and expand the flow of payments to their constituents and to relax the conditions under which patients are eligible for hospice benefits. Two such conditions are considered especially onerous: hospice coverage in Medicare requires both that patients renounce curative treatments and benefits and receive only comfort care and that they be certified by two physicians as having no more than six months to live.

Death, writes Blake Morrison (Reference Morrison1993: 203) is “a series of depletions”, so pinpointing remaining months of life is difficult both for physicians, who know the imprecision of such predictions in the face of clinical contingencies, and for patients, who may recoil from the finality of such estimates (however well-founded) and therefore avoid entering hospice care well past the ‘appropriate’ point (“People feel… that it's like signing a death certificate”, notes Diane E. Meier) (Amednews.Com, 2010). The Affordable Health Care Act of March 2010 nonetheless requires that state Medicaid and Children's Health Insurance programs cover ‘concurrent’ care, which combines curative and hospice services, for children with terminal illnesses and authorizes a three-year demonstration project of concurrent care for Medicare patients in 15 hospice-care sites.

Palliative care is a set of services, coverage for which its suppliers seek to secure and expand not only in but also beyond the hospice setting. These services have expanded impressively – between 2000 and 2006, the percentage of hospitals with 50 beds or more that report having a palliative care program roughly doubled, from 24% to 53% – and in the latter year the American Board of Medical Specialties approved palliative medicine as a new physician subspecialty (Center to Advance Palliative Care, 2008: 11, 35).

In seeking a firm financial foothold palliative teams face several challenges, however. Hospital executives, whose institutions make good money on intensive end-of-life care, may resist scaling it down, especially given that Medicare (and, as a rule, Medicaid) lack specific reimbursement categories for non-hospice-based palliative care in hospitals (Edens et al., Reference Edens, Harvey and Gilden2008). The ‘unbundled’ character of palliative services, which use ‘a patchwork of funding from multiple sources’ (TriCentral PC Toolkit, n.d.) and get combined with diverse types of care paid for in multiple settings, hospice and other, leaves payers fearful of opening their coffers to a flood of new claims for care rendered by team members other than physicians and in settings outside hospices. Palliative care providers, notes Diane Meier, are understandably nervous that they will be left out of the ‘inching toward capitation’ models now gaining favor (Meier et al., Reference Meier, Casarett, Von Gunten, Smith and Storey2010: 142). Gunnar Reference AlmgrenAlmgren (n.d.: 2, 4) laments that the ‘hospice paradigm’ of palliative care has ‘institutionalized a regrettable dichotomy’ that triggers ‘destructive competition’ between the two. Another expert, however, views these tensions as symptoms of a deeper analytical problem that policymakers have declined to tackle, namely, that “there is no existing inpatient model that encompasses the time-intensive interdisciplinary care” at issue, and so it will be necessary to create “an entirely new and innovative reimbursement structure” (R. Sean Morrison, quoted in Lubell, Reference Lubell2010).

Another challenge, which palliative care shares with hospices – and with nursing homes and home care agencies – inheres in the delivery of personal services to vulnerable clients by lower-paid workers in decentralized settings: monitoring, still less assuring, the safety and quality of care and the value of the money spent on it, is not easily done. Policymakers cope with these well-recognized problems by multiplying regulations that address accreditation, staff levels, training requirements, financial reporting and more in hopes of deterring ‘rogue hospices’ that put financial gain above the well-being of patients (Hospice Patients Alliance, n.d.). Oversight of caregivers’ performance and enforcement of regulatory constraints are complex and costly for regulators and regulatees, but such interventions seem to be the sole feasible way to avoid abuses.

On the other hand, well-intended regulations may work to the detriment of patients dying in nursing homes, where 20% of older Americans expire and from which 30% of all people who die in hospitals have been transferred. ‘Rehabilitative expectations’, cautions Sandra Johnson (Reference Johnson2005) may lead nursing home officials to transfer the gravely ill to hospitals (lest the homes be accused of failures of care), to supply insufficient management of pain (lest they be charged with overuse of drugs) and to impose tube feeding on the dying (lest it seem that they failed to do all they could to ‘promote health’). Crafting regulations that align (or at any rate do not misalign) the interests of dying patients and institutional caregivers is not simple.

As a recognized ‘institutional’ category of provider, hospices have made their (uneasy) peace with these formalities. By contrast, palliative caregivers, whose interests intersect imperfectly with those of hospices, hope that endorsement by prestigious organizations – the Institute of Medicine, the National Quality Forum [which designated palliative care as one of six national priorities for health care reform (Meier and Beresford, Reference Meier and Beresford2009: 594)], and the National Business Group on Health, for example – will win recognition that their work constitutes ‘the epitome of patient-centered care’ (Center to Advance Palliative Care, 2008: 9) and thereby confer legitimacy that can be translated into market leverage. Employers might persuade insurance plans, for example, to contract only with providers that deserve to be preferred because they offer acceptable palliative care benefits (Meier et al., Reference Meier, Casarett, Von Gunten, Smith and Storey2010: 143). In this legitimizing quest, however, palliative care (and hospices too) suffer from a ‘radioactive’ image acquired in the hullabaloo over death panels and from the inflaming of anxieties that health reformers are bent on rationing, on concocting ‘some mechanical governmental method… to determine how much of our scarce resources will be applied to [the patient's] situation” (Meier et al., Reference Meier, Casarett, Von Gunten, Smith and Storey2010: 141).

Aware that palliative care remains something of a niche player in the vast expanses of the health care system, its champions wonder why it has not come farther faster. Commitment to palliative care does not run deep in Congress, but this points to a deeper problem, namely, as a legislative aide explained, that “Congress isn't going to stick its neck out for something the public isn't clamoring for” (Meier and Beresford, Reference Meier and Beresford2009: 594). Pain management consumes only about 1% of National Institutes of Health research funds (Johnson, Reference Johnson2005: S39), and “no strong industry interests are aligned with good care for the end of life” (Lynn, Reference Lynn2005: S18). Political and industry champions are unlikely to emerge unless put under strong pressure by some sort of social movement, but cadres that might want to shoulder that burden – family caregivers, for example – have, in the nature of the case, more pressing claims on their energies.

Assisted suicide

Whereas policy for advance directives turns on the politics of the personal – devising structures and processes that allow the patient's preferences to be heard and honored in the conversational cacophony that often accompanies decisions about end-of-life care – policies for hospice and palliation turn on the politics of payment – designing strategies that enable suppliers of non-curative services to demonstrate the human and fiscal benefits of their work in terms that will secure for them a niche with public and private payers that is both reasonably stable and sufficiently generous to cover their costs and leave a margin for improvement and growth. Assisted suicide occupies a third realm, one dominated by the politics of principle – struggles over morality and rights, especially the alleged right of individuals to end their own lives by medical means should they elect to do so. The three revolve (to borrow terms from James Q. Wilson) on solidary, material and purposive considerations, respectively.

Advance directives and hospice and palliative care built momentum in the 1970s and surged subsequently. Assisted suicide (patients ending life by their own hand, with pharmaceutical assistance from physicians, which is distinct from euthanasia, in which a caregiver ‘administers’ a drug or other lethal agent) gained attention around the same time and was similarly propelled by policy entrepreneurs pressing the argument for death with dignity (in this case, ‘death counselor’ Jack Kevorkian and Derrick Humphry and his Hemlock Society, instead of Kubler-Ross and allies), but it has been adopted in the United States only in three states – Oregon, Washington and Montana.

Of the three, Oregon was the earliest adopter, having narrowly (51–49%) approved assisted suicide in a ballot initiative in 1994. Put on hold by a legal injunction, the state's electorate reaffirmed its support by a wider (60–40%) margin in 1997. It then faced myriad further legal challenges, including ones doggedly and repeatedly filed by the Justice Department in the administration of George W. Bush, and emerged from limbo only in 2006, when the US Supreme Court agreed by a vote of 6–3 that the matter was up to the states. In 2008, a state judge in Montana upheld the right of citizens of that state to choose assisted suicide, and in that same year, Oregon's neighbor to the north, Washington state, following a campaign in which former two-term Democratic governor Booth Gardner, who had come down with Parkinson's disease, threw personal support and substantial funding behind a ballot measure that approved the practice. Several other states, including Vermont, Michigan and New York, have entertained similar measures in their legislatures or on the ballot, but, battered by a phalanx of fervid and well-funded opponents at the state and national levels – the Catholic church, organized medicine, the disabled community, and right to life organizations – none has passed.

The assisted suicide ‘glass’ then is 6% full, and it is hard to say whether this is a little or a lot, or what it portends. That Oregon and Washington play Holland and Switzerland in the US federal system may derive from peculiarities in their political culture, especially that of Oregon. This state displays a distinctive, perhaps unique, taste for bioethical deliberation and for following arguments where they lead, however distasteful the end point may be. In the 1990s, Oregon adopted not only the nation's first assisted suicide statute but also its one and only program to ration (later ‘prioritize’) health services in its Medicaid program. The Oregon Plan for Medicaid emerged not from top–down directives but rather from (what were billed as) state-wide democratic deliberations within an explicitly utilitarian (‘greatest good of the greatest number’) ethical framework (on the Oregon plan see Oberlander et al., Reference Oberlander, Marmor and Jacobs2001). This plan resembled the assisted suicide law in the national, indeed international, notice it attracted, but the number of other US states (still less foreign States) emulating it (0) is even smaller than the number following suit on assisted suicide (2). Perhaps then assisted suicide will rest confined to the handful of states that substantially partake of whatever it is that distinctively animates Oregon's political culture.

The Supreme Court decision of 2006 is less than half a decade old, however, so arguably the spreading acceptance of advance directives, hospice and palliative care will encourage the citizenry to view assisted suicide as a corollary, not a reductio ad absurdum. Leaving the states secure in their constitutional control of the issue may encourage them to experiment in the proverbial laboratories of democracy in the US federal system with innovations that may diffuse, albeit slowly.

Another possibility is that the court's validation of state discretion and the exercise of it in favor of assisted suicide in three states will prompt the right to lifers and their allies on the political Right to redouble their efforts to stop assisted suicide in its tracks. They may well succeed, at least for a time, but that victory may ironically require them to scale back their political campaigns against advanced directives and palliative and hospice care because even very big and bilious movements lack the political capital to fight aggressively on several fronts at once.

End-of-life care as a work in progress

Releasing end-of-life care from institutional, technological, clinical and dogmatic imperatives is not easy in the United States, the medical cultural of which too often equates more care with better care, translates a macro commitment to conquer dread disease by means of research into micro determinations to hold the death of individuals at bay by all available clinical means, identifies service intensiveness with high quality, and overlays upon it all a conviction that the ways and means of ending individual lives (anyway those lives that are well insured) are somehow ‘society's’ business. But in the face of these obstacles, and contrary to both the frustrations of would-be plug pullers and the fears of death disempanelers, the United States in the last four decades has, haltingly and unevenly, put in place the policy rudiments of a formal–legal institutional framework that sustains a range of end-of-life decisions that conform better to the wishes of dying patients than did the decision-making patterns that prevailed before.

Advance directives, palliative care, hospices and (to a smaller extent) assisted suicide have advanced more or less in tandem and in response to broadly similar cultural and political dynamics. In each case, policy advocates and entrepreneurs determined to instill, or restore, personal dignity to dying pressed their case, acquired financial support and a measure of legitimation from philanthropies, sought new legislation and regulations at the state level, pushed on toward federal reinforcement (e.g. Medicare payment for hospices in 1983 and the Patient Self-Determination Act of 1990), saw the innovations in question become increasingly institutionalized, built organizational alliances and expanded their strategies for lobbying and social marketing.

Policy coordination – the design of seamless, integrated end-of-life options – remains (and is likely to remain) elusive for at least three reasons: because much of the action is at the state level, which encourages policy improvisation, but does not guarantee diffusion of innovations or their uptake at the national level; because the politics of the personal, of payment, and of principle tend to galvanize different actors and organizations to coalesce and contend around distinct objectives (boosting the penetration of advance directives in the population, winning higher payment on more reasonable terms for palliation and hospice care, persuading state political leaders to accept an assisted suicide option); and because tensions persist within each realm (the relative merits of living wills vs durable POA, the unevenness of the playing fields for palliative care and hospices, how best to deploy Oregon's data to strengthen the case for assisted suicide).

All the same, the groundwork laid seems to leave the United States well within cross-national ‘norms’ for end-of-life care of good quality. A recent report by the Economist Intelligence Unit (2010: 11–15) lists the United States as tied with Canada for 9th place out of 40 nations on its ‘overall’ quality of death index, 9th on ‘basic end-of-life healthcare environment’, 8th on availability of end-of-life care, 7th for the quality of end-of-life care, and in cluster number 4 (on a scale in which 5 is excellent and 1 is poor) on measures of availability of pain killers. On the cost of end-of-life care the United States, unsurprisingly, performs less well (18 of 40).