Voiceless and Vulnerable: Dementia Patients Without Surrogates in an Era of Capitation

In 1990, the population prevalence of dementia in the United States was estimated at 10.2% of the 65 and older age group and 40% of people aged 85 years and older.1 The number of persons more than 85 years of age, the group most likely to suffer from dementing illness, is projected to quadruple over the next 40 years. As a result, in the year 2040, the prevalence estimate for dementia lies somewhere between 7 and 12 million persons, almost half of whom will have advanced disease and will require complete custodial care, often for years, before death.2 Medical decisions for dementia patients are usually made by family members acting in consultation with clinicians. Ideally these decisions are based on knowledge of the patient's previously expressed wishes, such as are found in advance directives, or on knowledge of the patient's wishes gained from specific conversations on the topic occurring before the onset of significant cognitive impairment. This kind of evidence of patient preference does not exist for the vast majority of dementia patients, either because the onset of dementia predated the widespread availability of advance directives or because physicians, patients, and family members failed to engage in the necessary discussion about preferences before the patient lost the capacity to participate. In the absence of such information, surrogate decision-makers are generally advised to decide as they think the patient would have under these circumstances or to decide based on what they think would constitute the patient's best interests. In most states, surrogate decision-making laws support such a family centered model of decision-making for patients who lack decisional capacity.3 Though there may be some risks to family centered surrogate decision-making for patients without decisional capacity, studies suggest that this is the approach preferred by patients and is that most likely to preserve both their autonomy and their best interests. But what about dementia patients who have no one to speak for them when decisions must be made on their behalf? Very little is known about the numbers of dementia patients in the United States who either have no one to serve as a surrogate decision-maker or whose potential surrogates are incapable, unavailable, or otherwise not functional when decisions must be made. A MEDLINE search and extensive personal communication with colleagues in the field revealed almost no data on the prevalence of this problem, a statement in itself of the lack of attention to this extremely vulnerable, but voiceless, population. In a study by Fader et al., 45% of next of kin listed for nursing home residents could not be reached for participation in major decisions about their relative's resuscitation status, suggesting that a significant number of long-term care residents are functionally without a surrogate.4 A 1994 survey on treatment decision-making policies and procedures for decisionally impaired residents of nursing homes in New York State found that 27 to 37% of respondents believed that major medical decisions “sometimes or often” had to be made for patients without families.5 In an ongoing randomized controlled trial of palliative approaches to the care of hospitalized patients with advanced dementia conducted at Mount Sinai Hospital in New York City,6 45% of otherwise eligible, severely demented, subjects could not be randomized because they either had no surrogate or their surrogate was unable to engage in any deliberative process on their behalf. Medical decisions for these acutely ill patients with advanced dementia were made by their attending physician, presumably based on the attending physician's conception of the patient's best interest. Similar reversion to physician-based decisions are presumed to occur in long-term care settings when no surrogates are available.7 See also p 369 Inadequate though this data is in estimating the prevalence of dementia patients without surrogates, it suggests that the problem is not rare and is likely to increase with the growth in the population at risk. Whether the default method of medical decision-making by the physicians of cognitively impaired persons without surrogates actually does serve their best interests remains an unresolved question. Multiple studies suggest that physicians tend to make allocation decisions based on their perception of patient social worth, leading to clear inequities based on characteristics such as race, age, and gender.8-10 At present there is no safeguard or consent mechanism for treatment or for decisions to forgo treatment in this population short of recourse to the courts, a requirement that may lead to delays in needed therapies or continuation of burdensome and inappropriate life supports in patients suffering from the terminal stages of dementing illnesses. The fee-for-service model for health care delivery may have served as a partial safeguard on any tendency to withhold services or undertreat. It is equally likely that fee for service incentives may have reinforced a tendency to overtreat, or to use or continue life-prolonging technological interventions such as tube feeding or antibiotics, even for patients in the most advanced and terminal stages of Alzheimer's Disease. The question that faces us now is what impact the new incentives accompanying managed care and full risk capitation will have on physician's and institution's view of the patient's best interests. If financial interest of doctors and healthcare facilities is best served by doing less, how will this uniquely vulnerable group be protected from inappropriate undertreatment? Although HMOs have been demonstrated to reduce costs and to maintain or improve quality of care to the general population,11 data is now beginning to emerge about the impact of managed care on health outcomes that suggests that the fiscal restraint imposed by this system may adversely affect high risk populations such as the poor, the chronically ill, and the aged. In a recent 4-year observational study12 comparing physical and mental health outcomes in chronically ill, poor, older adults treated either in HMO or fee-for-service systems, the likelihood of decline in physical health in an HMO was more than double that observed in the fee-for-service setting (68% in HMO vs 27% in fee-for-service, P < .001). Another study of healthcare expenditures and utilization in a frail older population in San Diego13, 14 demonstrated a fourfold higher rate of hospital readmission, a 3.5-fold higher rate of preventable hospital readmissions, and fewer home health visits in HMO enrollees compared with Medicare fee-for-service and Medicare-Medicaid patients. Several recent studies15, 16 found lower resource use and costs in seriously ill older patients compared with otherwise similar younger patients, suggesting that some allocation of resources based on age per se is already occurring, even in fee-for-service settings. The managed care share of the Medicare market has doubled since 1985 and is projected to grow rapidly in the near future,17 suggesting that pressures for undertreatment attributable both to heightened cost containment pressures and the already occurring age-based rationing of resources described in recent studies will be an unavoidable reality of medical care for patients with advanced dementia. How can this population be protected from neglect or undertreatment? Many have argued that the ideal solution would include more use of advance directives executed before the onset of clinically significant incapacity to guide future health care decisions. But reliance on an advance directive to promote patient autonomy is an illusion when the fiscal pressures of the health care environment permit no real choices and focus increasingly on enhancing the bottom line. Because advance evidence of patient wishes is still uncommon and is, in any case, unlikely to be a force powerful enough to resist social forces pressing for undertreatment, an advocacy mechanism must be created to ensure consistent adherence to appropriate standards of care and prevent undertreatment out of expediency, perverse financial incentives, or neglect and overtreatment out of lack of any mechanism to permit discontinuation of life supports. A recent position statement from the Ethics Committee of the AGS18 proposed local multidisciplinary team decisionmaking on behalf of dementia patients without surrogates combined with routine guardianship proceedings and ethics committee mediation for resolution of conflicts. In an associated editorial, Pearlman argued that physicians tend to err on the side of undertreatment on behalf of incapacitated patients19 and that since ethics committee processes are neither standardized nor regulated, many are unable to participate in the rigorous ethical analysis, patient advocacy, or conflict resolution expected of them. Further, the imposition of the strong physician and hospital cost containment incentives associated with managed care will place physicians and institution-based ethics committees in a position characterized by a stark and irreducible conflict of interest: their own or their colleague's or their employer's financial survival versus the often costly best interests of an individual patient.20 Precisely for this reason, Callahan21 has argued for an open and public policy of age-based rationing, one that would shift the background assumptions from open-ended entitlement to clearly defined and fairly applied limits on expensive care at the end of life. At the very least, such a policy would aim for an equitable set of limits on care rather than allowing a disproportionate share of the burden to fall on the most vulnerable groups — those in poverty, those without family advocates, and those who are cognitively impaired. Given the inevitable collision of the vulnerability of cognitively impaired persons without surrogates with overwhelming cost containment pressures, what should an ideal policy look like? The ideal policy would be locally based, responsive to the rapidly changing realities of clinical practice, and independent of any direct financial interest in the outcome of the decision. The degree of regulation and oversight should be proportionate to the consequences of the decision: higher for decisions to forego life-sustaining therapy or begin long-term tube feeding and lower for more routine treatment decisions such as initiation of an antibiotic. Because of costs, delays, and bureaucratic barriers to their efficient use, recourse to the adversarial environment of the courts for appointment of guardians is rarely sought and should not be mandated unless the process can be greatly simplified and expedited. In this issue of the Journal, Miller et al.22 argue for a facility-based model of physician decision-making for the population of dementia patients without surrogates. They point out that appointment of a guardian currently requires strictly regulated, time consuming, costly, and highly bureaucratic judicial proceedings in most states; that guardians are scarce, particularly for patients without assets; and that the guardians themselves are neither trained nor monitored and have not been shown to be effective advocates for the mentally disabled. Because of these difficulties, Miller at al. propose reliance on the patient's primary physician with a concurring second opinion for decisions about major medical treatments; decisions to forgo or withdraw life-sustaining treatments would be subject to a higher level of regulation and require adherence to a set of clinical standards (treatment poses an excessive burden, patient is terminally ill or permanently unconscious) determined by the attending physician, a concurring second opinion, and prospective review and approval by a multidisciplinary committee (including community and/or family representation) within the facility. Although confidential, these deliberations could be subject to retrospective review by legal or policy bodies. This proposal has the virtue of simplicity, feasibility, and local control of local decisions. It cannot, however, protect the patient from inappropriate undertreatment when the very individuals and facilities charged with the decision stand to gain financially from deciding not to treat. Although nursing homes are not currently subject to capitation, there is no reason to believe they will not ultimately be included in the services provided by full risk plans. Therefore, the model of local physician and committee decision-making will have to be balanced and supplemented by routine appointment of guardians for dementia patients without surrogates. Given the risks of failing to improve access to guardians, it should be possible to simplify and reduce the costs of the judicial requirements for guardianship proceedings, to create training and monitoring bodies for guardians, and to establish public sources of funding both to compensate guardians and to support the programs that will ensure their competence as advocates for isolated patients without decisional capacity. Perhaps the insurers and the institutions responsible for the care of these patients should be mandated to contribute to such a fund. In any case, the large numbers of patients who will be affected, their extremely high risk of expedient under-treatment, and the obligation of society to protect its most vulnerable members requires urgent legislative and judicial action to facilitate a functional and efficient guardianship process. In combination with local judgments made by the patient's professional caregivers, and backed by second opinions and facility-based committee review, these safeguards should begin to provide the protections quite obviously needed. Geriatricians should work with state legislatures and local bar association groups to promote clinically feasible and responsible policies. Easier access to quality guardianship combined with facility based decision-making is the most reasonable alternative before us.