Toward Revising the Ethical Boundaries of Research With Noncompetent Subjects

Over the past few years, several themes have emerged in the field of dementia. The consensus among experts that dementia research of all types is critical has strengthened as the baby boom generation enters its seventh decade, its members joining the growing population of elderly at risk for neurodegenerative disorders. At the same time, scientific progress in the prevention and treatment of dementia has been disappointingly modest (1) .

Another continuing trend is the widespread support for patients’ and study subjects’ rights that emphasizes both the primacy of autonomous decision making and the need to protect persons, especially the elderly, who may have diminished cognitive capacity. Indeed, the federal regulations governing human subject research, “the common rule,” require that vulnerable persons be afforded “additional safeguards” (45 CFR 46.111). Furthermore, the public perception of the research enterprise as beneficent has recently been shaken by reports of investigator-pharmaceutical financial alliances that appear to threaten scientific objectivity (2) . The challenge for investigators is to develop innovative research methods that can facilitate enrollment of subjects with cognitive difficulties while respecting core ethical principles and acknowledging the public skepticism about the medical-industrial complex.

Much has been written concerning the ethical and practical difficulties of two research scenarios—enrolling competent but at-risk subjects into future research on dementia that has not yet been developed and enrolling already cognitively impaired subjects into current biomedical research (3 , 4) . A decade ago, APA issued an official statement warning that “the purpose of attending to assessment of the capacities of potential research subjects is to ensure that only subjects capable of giving consent are asked to make decisions about research participation. This protects both subjects’ interests and the integrity of the research process” (5) .

In instances where cognitive impairment precludes obtaining consent for diagnostic and treatment interventions, there is an accepted mechanism for soliciting the approval of a health care proxy. However, proxy consent for research purposes is permissible only in the most narrow of circumstances, where the risks posed by nontherapeutic investigations (those without potential benefit) are no more than minimal (such as is ordinarily encountered in daily life) and are justified by the importance of the knowledge being sought in the study (6) .

This issue of the Journal features an important article by Karlawish and colleagues at the University of Pennsylvania that provides quantitative data on attitudes about dementia research in a community population of elderly persons (7) . Of more than 500 participants in the study, 92% would permit being enrolled in a study involving blood drawing and 75% would permit a lumbar puncture either by granting advance consent or by granting their proxy leeway over advance consent. Collecting such data about preferences is a necessary but preliminary step in developing models of ethical research that minimize adversarial conflicts of interest among investigators, subjects, patient advocates, and funding agencies, including pharmaceutical companies. The opinions of these potential future subjects should help ensure the acceptability of novel consent procedures for persons who are no longer capable of appreciating the risks and benefits of participating in medical research.

The solution of allowing for “leeway consent” from a health care proxy for dementia research is appealing but not well defined. The appropriate standard for decision making by a proxy has been controversial despite U.S. Supreme Court guidance in the 1990 case of Cruzan v. Director, Missouri Department of Health(8) . Nancy Cruzan was a young woman who went into a vegetative state as a result of a motor vehicle accident. Cruzan’s parents asked to terminate their daughter’s artificial feeding and hydration. Missouri law prohibited the parents’ request because they had failed to show by clear and convincing evidence that their daughter would have wished to have her life-sustaining treatment withdrawn. The parents appealed, and the case eventually reached the U.S. Supreme Court, which sided with Missouri in a five-to-four decision. This precedent would direct the proxy to grant substitute consent only when there is strong reason to believe that consent is what the subject would have wanted. In the absence of such evidence, no consent would be given.

There is an alternative standard that would depend on what would be surmised as the potential subject’s best interests. Karlawish et al. found that individuals showed a willingness to grant their proxy decision maker leeway over future choices regarding consent to participate in research projects. Such willingness was most strongly associated with a favorable attitude about biomedical research. These results would suggest that research that does not offer any specific benefit and involves minimal risk of harm is generally perceived as acceptable by the elderly. An inference would follow that a particular individual who subsequently becomes incompetent to consent to participation in research would also find enrollment in the study to be consonant with his or her best interests, regardless of whether such a preference had previously been expressed.

Over the past few years, significant changes in the way biomedical research is being conducted in the United States have included the migration of research from university and voluntary hospitals to proprietary clinical trial companies, larger reimbursements for study subjects, and greater transparency of study protocols and results. The scientific community should be optimistic that the domain of research that would permit proxy consent can be expanded beyond the narrow confines of minimal-risk studies.

The hypothetical questions posed by Karlawish et al. referenced hypothetical low-risk/low-potential-gain studies. How does a proxy decision maker choose in research that is of higher risk but with a higher potential for reward?

The two-decade-old Appelbaum-Grisso model for consent for medical treatment consists of the abilities to communicate a choice, to understand the relevant information, to appreciate the medical consequences of the situation, and to reason about treatment choices (9 , 10) . Unlike other models of consent, these criteria use the word “appreciate,” which signifies the addition of an emotional valence to the level of comprehension. While the task of a proxy decision maker may be easier when considering the issues on a purely objective level, the affective dimension becomes one of several complicating factors. Should the feelings of family members be given preference? What if the proxy consents, but the noncompetent subject actively resists participation? Should the proxy be empowered without advance directive guidance to permit sequential MRI and CSF studies, DNA analysis, and postmortem tissue collection and brain banking—all of which may produce greater research yields? These are but a few examples of the potential conundrums that can arise.

Progress regarding appropriate moral boundaries of geropsychiatric research depends on accumulating relevant quantitative data. The article by Karlawish et al. is a noteworthy early contribution to the evidence base that will be necessary to expand opportunities for the participation of elderly subjects in studies of Alzheimer’s disease and other neurodegenerative disorders.