In Oregon, the public discussion leading up to the 1994 and 1997 votes to legalize physician-assisted suicide was intense and polarized. During both campaigns, media advertisements obscured the substantial common ground between the two sides. The 1994 vote, however, did serve as a wake-up call to medicine and fueled progress in improving the quality of end-of-life care [4]. The resulting major improvements in Oregon's end-of-life care are the work of hundreds of persons and a variety of programs and health systems from all parts of the "assisted suicide" spectrum.

Quill and colleagues [1] highlight the consequences of inadequate planning for end-of-life care by reviewing the results of the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT), which confirm that most Americans die in acute-care hospitals with high utilization rates for invasive medical technology [3]. These procedures and treatments are sometimes performed against the patient's expressed wishes [1, 3]. National interest is therefore keen in developing and implementing "signalling" mechanisms, including such things as standardized order forms and bracelets, that can help ensure respect for patient wishes to limit life-sustaining treatments [5]. Over a 5-year period, a 40-person task force developed, implemented, and pilot-tested such a physician order form, which recorded patient wishes to have or limit life-sustaining treatment, for statewide use in Oregon [6, 7]. The bright pink form, known as Physician Orders for Life-Sustaining Treatment (POLST), is respected across settings (for example, in the nursing home, the ambulance, and the acute-care hospital) [7]. Most nursing homes and hospice programs are now using the POLST form, and preliminary data confirm that these orders are rarely ignored. The POLST form functions within the context of Oregon's liberal advance directives law and would require modification for use in some other states (for example, those without surrogacy provisions).

Substantial geographic variation occurs in the in-hospital mortality rate [8]. Of all 50 states, Oregon has the lowest in-hospital mortality rate. Surveys show that when given the choice, most persons prefer to die at home or in a home-like setting [9]. Oregon's low rate of in-hospital deaths suggests the presence of a high level of advanced care planning and respect for patient wishes. In 1996, only 31% of Oregonians died in acute-care hospitals [10]. Medicare data show that the rate among persons older than 65 years of age was even lower. Portland, Oregon, where the POLST form was first tested, has the lowest in-hospital death rate of the 306 ZIP code areas examined by Weinberg and colleagues [8]. Only 22% of Medicare patients die in acute-care hospitals in Portland; this can be compared with a high of 51% in Newark, New Jersey. The rate of in-patient deaths in many eastern U.S. cities is double that in Oregon [8]. What percentage of deaths should ideally be in-hospital deaths? We really do not know; the correct number clearly is not zero, however, and depends partly on whether resources are adequate to ensure comfort and clinical support in an out-of-hospital setting.

Quill and colleagues call for the increased use of hospice and palliative care in assuring comfort during terminal illness. Oregon has shown increased attention to comfort measures in two ways. First, the rate of admission to hospice increased 20% in 1995 and continues to increase [4]. Preliminary estimates suggest that Oregon has moved from fifth to third in the rate of hospice admissions over the past 2 years. Of Oregonians who die each year, 29% are doing so with hospice support; the national Figure is only 17% (Jackson A. Personal communication).

Second, Quill and colleagues also refer to the tremendous amount of suffering experienced by dying patients across the United States [1]; SUPPORT found that 50% of those who died had moderate to severe pain at least half of the time during their last 3 days of life [3]. In most states, morphine and other medications to control moderate to severe pain and suffering are underprescribed. Oregon currently leads the United States in the medical use of morphine, as measured in milligrams per 100 000 persons. From 1994 through 1996, Oregon increased its medical use of morphine by 70% [11]. Oregon is fortunate to have fewer impediments to appropriate prescribing practices than some other states (for example, we do not have triplicate forms).

Quill and colleagues [1] acknowledge, moreover, that 40 million U.S. citizens are uninsured and therefore have limited access to a wide array of health services, including hospice care for terminally ill persons. Oregon has made significant progress in this area. Through the Oregon Health Plan, the state now provides benefits to most of its previously uninsured, terminally ill citizens. Since 1994, the percentage of uninsured persons enrolled in Oregon hospice programs has decreased from 15% to 2% (Jackson A. Personal communication). Although Oregon has significantly reduced its numbers of uninsured persons, it still must face the sad reality brought forward by Quill and colleagues: Some dying Americans do not have the financial means to ensure access to hospice and other comfort measures [12]. The experience of those with and those without health insurance are not the same [13]. Lack of funding reduces access to comfort measures. Terminally ill persons without health insurance are more likely to be referred to hospice late and in crisis, if at all (Jackson A. Personal communication).

Society has achieved consensus on some aspects of end-of-life care. We agree that terminally ill patients have the right to refuse life-sustaining treatments, and we have developed advance directives to protect this right if the patient becomes incapacitated. Although we sometimes fail to listen to patients and families who want treatment stopped, the courts and ethics consultants stand ready to defend this right. The U.S. Supreme Court did not find a constitutional right to physician-assisted suicide, but the justices hinted at a right to comfort measures when they cautioned states not to create procedural barriers to palliative care [14].

Both sides in the assisted suicide debate agree that lack of funding reduces choices and options for uninsured persons. These persons face the prospect of impoverishing their families in order to receive care in life's final months. Families experiencing a greater financial burden are more likely to agree to the withdrawal of a loved one's life support [15]. To avoid leaving a young family in poverty, some uninsured, terminally ill persons may be more likely to pursue assisted suicide than if they had adequate insurance coverage. In this respect, the option of physician-assisted suicide remains essentially a coercive choice for persons without insurance coverage for hospice or comfort measures.

We estimate that it will cost $1.4 million per year to provide universal access to hospice for terminally ill Oregonians who remain uninsured [12]. Ironically, those for and against Oregon's assisted suicide initiative spent a total of $4.6 million in advertising for the intensely debated November 1997 election to retain or eliminate Oregon's assisted suicide law. That is enough money to fund access to hospice for every dying Oregonian for 3 years. If so much money can be found to fight battles over differing moral views, it seems reasonable to expect that, as a society, we can agree to fund access to high-level end-of-life care, including hospice care for all terminally ill Americans, thereby ensuring a true choice about end-of-life treatment.