8Research Ethics

Publication Details

It became evident early in the panel's deliberations that investigators are uncertain about the ethical requirements governing research on elder mistreatment. They are particularly concerned about two issues: (1) Under what circumstances is it ethically necessary to exclude elderly persons from participation in research on the ground that they are incapable of giving informed consent? (2) How should investigators respond to evidence of mistreatment elicited during the study, and what should participants be told about this problem in advance?

It also became apparent to the panel that Institutional Review Boards (IRBs) have embraced varying approaches to these problems, and that unduly restrictive IRB positions could impede important advances in the understanding of elder mistreatment. In reviewing proposed research protocols in the current regulatory context, IRBs exercise a great deal of discretion within the general parameters set by the federal research regulations (i.e., the so-called Common Rule) (U.S. Department of Health and Human Services, 2001a) that govern all federally funded research, and the regulations of the U.S. Food and Drug Administration (1999) that govern clinical trials of drugs and devices. To help provide better guidance to investigators and IRBs in preparing and reviewing elder mistreatment research protocols, the panel commissioned a paper by Rebecca Dresser, a leading authority on research ethics. Her paper (this volume) surveys the issues that tend to recur in research on elder mistreatment and offers many helpful suggestions about how to respond to them.


As Dresser (this volume) indicates, many of the concerns in elder mistreatment research relate to the assessment of decision-making capacity and the proper responses to situations in which potential participants' decisionmaking capacity is impaired. Taking into account that elderly persons who are most vulnerable to mistreatment are likely to be decisionally impaired, and that decisional impairment is often embedded in the statutory definition of mistreatment, the opportunity to elicit information from cognitively impaired individuals is essential in almost every study of elder mistreatment. The problem is by no means limited to elder mistreatment research; similar observations could be made about many other types of research focusing on older subjects or on people with mental retardation or schizophrenia. Not surprisingly, ethical issues relating to consent by cognitively impaired subjects and surrogate decision making have come to center stage in recent years and have been addressed in a rapidly developing literature (Berg, 1996; Bonnie, 1997; Dresser, 1999; National Bioethics Advisory Commission, 1998). Although the Dresser paper provides a thorough review of these issues, several points are emphasized here.

First, aging research suggests that most older adults are cognitively intact and should be regarded as presumptively able to make informed, voluntary decisions about research participation; merely being older than age 65 or 75 does not warrant special screening procedures or other protections. Heightened safeguards should be considered when the research involves greater than minimal risk and the study population is especially likely to include persons with decisional impairments or persons who are especially vulnerable to pressure or influence.

Second, a diagnosis of dementia is not congruent with decisional incapacity (Marson et al., 1995). Instead, an assessment of decisional capacity requires a highly contextualized judgment concerning a particular person's ability to perform ethically relevant decision-making tasks in relation to a particular study. Significant advances have recently been made in conceptualizing, operationalizing, and measuring these decision-making abilities. For example, according to one widely used approach, the ethically relevant abilities in making informed decisions about treatment or research participation are understanding of risks and benefits and other ethically relevant information, appreciation of the relevance of that information in one's own situation, reasoning or processing the information logically, and ability to make a stable choice among alternatives (Appelbaum and Grisso, 1988; Grisso and Appelbaum, 1998). Instruments are being developed to help researchers assess and document participants' decision-making abilities, both at the initiation of the study and, if needed, on a continuing basis during the study (Kim et al., 2001; Marson et al., 1995).

Third, it is understood that the decision-making abilities of many older people with dementia or other mental disorders will be impaired to some extent. However, mere impairment does not amount to incapacity. Whether a person's impairment (in understanding, appreciation, or reasoning) is substantial enough to preclude informed consent for a particular study depends on the complexity of the decision-making task for that study, and on its risk-benefit profile (Grisso and Appelbaum, 1998; American Psychiatric Association, 1998; National Bioethics Advisory Commission, 1998). A minimum threshold of capacity must be satisfied for all studies in which the participant is exposed to any risk, but the level of capacity needed above this floor will depend on the characteristics of the particular study. Moreover, impairments that would preclude valid consent in studies with significant risk are ethically immaterial in studies involving less than minimal risk (National Bioethics Advisory Commission, 2001). Similarly, additional safeguards that may be needed to ensure proper consent in studies involving significant risk are not necessary in studies involving lower risk. Risk stratification of this kind is a prominent and essential feature of sensible ethical review.

Fourth, even if an elderly person lacks the capacity to give informed consent for the particular study, his or her participation may be authorized by a surrogate decision maker as long as the subject is adequately protected from harm and the IRB finds that all the other criteria specified in the Common Rule have been met. The Common Rule, however, provides very little guidance to IRBs, within these parameters, about the conditions under which surrogates should be permitted to authorize participation for decisionally incapable subjects in lieu of excluding them from participation altogether (Bonnie, 1997; Dresser, 2001; National Bioethics Advisory Commission, 1998). This lack of guidance has led to inconsistency among IRBs and frustration among investigators. However, recent reports by expert bodies have begun to provide such guidance and to assemble best practices (National Bioethics Advisory Commission, 1998; American Psychiatric Association, 1998). The important point for our purposes is that the Common Rule provides ample flexibility to IRBs to allow important research on elder mistreatment to go forward on the basis of surrogate consent, particularly if it involves less than minimal risk.

Finally, another concern is the availability of a suitable surrogate. The Common Rule refers generally to individuals “authorized under applicable law to consent on behalf of a prospective subject's participation.” As Dresser (this volume) notes, however, many states lack clear rules in this area. To the extent that impediments to research arise from ambiguities or constraints in state law governing surrogate decision making, these are generic problems in research that, in the end, depend on clarification of state law. The panel urges states to extend their rules governing surrogate decision making for health care decisions to the context of research (see Hoffman and Schwartz, 1998).

It is possible that the most suitable surrogate would be a family member whose own improper conduct would be exposed by the study. Under these circumstances, involving possible conflicts of interest between the decision maker and the potential research subject, other suitable surrogates should be identified. As Dresser (this volume) notes, analogous problems in child mistreatment research have been addressed in the subpart of the Common Rule prescribing additional protections for children participating in research (U.S. Department of Health and Human Services, 2001a). Overall, it should be emphasized that, in the absence of impediments rooted in state law, the IRBs have a great deal of flexibility in permitting research to go forward with decisionally incapable participants as long as a suitable surrogate authorizes participation and the participants are protected from harm.


Another major area of ethical uncertainty with particular salience in elder mistreatment research concerns ethically permissible responses to evidence of mistreatment that may be elicited during the study (and the associated issue of what potential subjects should be told about this possibility during the consent process). Deciding how the investigator should respond to evidence of mistreatment implicates the investigator's most fundamental ethical duties in human research—the duty to protect participants from harm and the duty to respect their autonomy as persons—and also exposes possible conflicts between the interests of victims of mistreatment (in being protected) and suspected perpetrators of mistreatment (in avoiding possible punitive interventions). Because both the possible perpetrator of mistreatment and the possible victim could be regarded as subjects of (or participants in) the research, the investigator may owe both of them a duty to avoid harm.

If these puzzles were not enough to create ethical complexity, the possibility that state laws may require researchers to disclose evidence of mistreatment to official agencies highlights a possible incongruity between state reporting policies (requiring breaches of confidentiality) and federal research policy (emphasizing the need to protect confidentiality). It can readily be seen why IRBs would be perplexed by these problems (which also arise in research that could elicit evidence of child mistreatment) and why some IRBs might end up embracing policies that effectively preclude otherwise meritorious studies.

Because the goals and methods of elder mistreatment research vary so widely, it is neither possible nor desirable to formulate generally applicable policies or rules about how to resolve these problems. The most sensible approach is to avoid hard-and-fast rules altogether and to permit investigators and IRBs to tailor their responses to the settings and protocols of the particular studies. With this caveat in mind, the panel recommends that the National Institute on Aging, in collaboration with the Office for Human Research Protections and other sponsors of elder mistreatment research, undertake a consensus project to develop ethical guidelines for responding to evidence of elder mistreatment in various types of research settings (e.g., telephone surveys, in-person interviews in clinical settings). As part of this process, representative samples or groups of elder persons, caregivers, and other family members should be surveyed to ascertain their attitudes on the relative importance of the competing values at stake and the appropriate responses to the dilemmas regarding disclosure of information bearing on possible mistreatment.

In the absence of more specific guidance emerging from such a consensus process, the panel suggests that investigators undertake to obtain community guidance about some of the ethical difficulties raised by their particular protocols (see Levine, 1986; Melton et al., 1988, Strauss et al., 2001). Whenever feasible, investigators should consult representative members of the populations being studied (elder persons and caregivers, nursing home residents and staff, etc.) to ascertain their perspectives and preferences regarding the proper responses to evidence of mistreatment (and the related ethical issues raised by the proposed research) and should take this information into account in developing the protocol.

It appears that the biggest impediment to sensible and consistent resolution of the ethical issues raised by elder mistreatment research is the concern that state law requires researchers to report possible cases of mistreatment to public agencies. Under some circumstances involving clear danger to the elderly subject, researchers may certainly feel ethically bound to take preventive action, and sometimes this may require them to call adult protective services authorities. However, evidence of mistreatment that may be elicited during the interview is often not indicative of current danger, and there are many other ways in which ethically sensitive and caring researchers can respond to such evidence. It should also be recognized that reporting (and triggering an adult protective services investigation) can have disruptive effects on the life of an elderly subject and may not be in his or her best interests. For these reasons, in the panel's view, evidence of mistreatment elicited during research should not trigger a mandatory obligation to report. Whatever may be thought of mandated reporting in ordinary practice settings (see Chapter 7), it is too blunt a response in the research context. Instead, whether the researchers should voluntarily take protective action and, if so, whether such action should include reports to public agencies should be determined by researchers and IRBs in the context of the goals, setting, and methods of the particular study. Elder mistreatment reporting statutes should be amended to exempt researchers from their mandatory requirements.

The requirements otherwise imposed by state reporting statutes can be overridden by a certificate of confidentiality issued by the National Institutes of Health (NIH) in accordance with federal law. In the panel's view, NIH should issue certificates of confidentiality designed to insulate elder mistreatment researchers from any legal obligation to disclose possible cases of mistreatment that otherwise may arise under state law, including tort “duty to protect” obligations as well as reporting statutes. Issuance of these certificates should be predicated on the assumption that IRBs will carefully scrutinize the protocols to ensure that participants' interests are being protected from harm and that, under appropriate circumstances, IRBs will permit investigators to take voluntary steps to protect subjects in danger. (This point is discussed in greater depth below.)

The panel is confident that the federally issued certificate of confidentiality supersedes otherwise mandatory reporting requirements under state law. Pursuant to §301(d) of the Public Health Service Act, 42 U.S.C. § 241(d):

The Secretary may authorize persons engaged in biomedical, behavioral, clinical, or other research (including research on mental health, including research on the use and effect of alcohol and other psychoactive drugs) to protect the privacy of individuals who are the subject of such research by withholding from all persons not connected with the conduct of such research the names or other identifying characteristics of such individuals. Persons so authorized to protect the privacy of such individuals may not be compelled in any Federal, State, or local civil, criminal, administrative, legislative, or other proceedings to identify such individuals.

Although some researchers have wondered whether a person is being “compelled” to “identify” their subjects by a statutory reporting obligation for child abuse or elder mistreatment (see Gray et al., 1995), it is clear that a statutory reporting obligation, with penalties for failure to report, is a form of compulsion, and that a report of the perpetrator inevitably identifies the victim. It is also clear that a certificate of confidentiality issued pursuant to this statute will supersede the state's reporting obligation under the supremacy clause of the Constitution (under which federal law is the “supreme law of the land”).

Although the issue has never been litigated in federal court, the certificate's authority has been recognized by the highest court of New York in People v. Newman, 345 N.Y.S. 2d (1973) (certificate issued pursuant to an equivalent provision in the Comprehensive Drug Abuse Prevention and Control Act of 1970 protected Dr. Newman from a subpoena requiring disclosure of pictures of patients receiving methadone).1 It should also be noted that regulations issued by the U.S. Department of Health and Human Services (2001b), to which courts would give considerable weight in interpreting the federal statute, do not include state reporting obligations in the list of disclosures permitted by a certificate of confidentiality. Finally, NIH guidelines pertaining to confidentiality certificates include the following provisions:

Background: Confidentiality Certificates are issued by NIH Institutes pursuant to Section 301 (d) of the Public Health Services Act (42 U.S.C. Section 241(d)) to afford special privacy protection to research subjects. A Certificate helps the researcher avoid compelled “involuntary disclosure” (e.g. subpoenas) of identifying information about a research subject. It does not prevent voluntary disclosures such as disclosure to protect the subject or others from serious harm, as in cases of child abuse. Also, a researcher may not rely on a Certificate to withhold data if the subject consents to the disclosure.

Informing subjects about certificate: When a researcher obtains a Confidentiality Certificate, the subjects must be told about the protections afforded by the Certificate, and any exceptions to that protection. This information is usually included in an “informed consent.”

Need to adapt examples: Research subjects vary widely in their cultural and educational backgrounds. The language used should cover the basic points—privacy protection means that the subject will not be identified as participating in the study, unless the subject consents, or a disclosure is made to protect the subject or another from serious harm. Researchers may adapt the language to the special needs of their clientele, and to the subject matter of the study.

Researchers should also review the language about confidentiality which is routinely included in consent forms to be sure that it is consistent with Confidentiality Certificate protections. For example, consent forms sometimes refer to state law reporting requirements. However, HHS General Counsel advises that such a disclosure would be voluntary, even though (otherwise) required by State law, because the Certificate protects the researcher from the compulsion of that law. Thus, note that the examples given simply state the circumstances in which disclosures would be made. [emphasis added]

Thus, NIH policy clearly supports the contention that certificates exempt researchers from the obligation to report elder abuse under state reporting statutes.

Assuming that disclosure to public agencies is not mandated by law, the next question is whether the research protocol itself should require or permit such disclosure, and, if so, under what circumstances. In the panel's view, direct disclosure to public authorities should never be required by the protocol, just as it should not be required by law. The better approach is to ask the mistreated person (or the perpetrator, if that person has made the disclosure) whether he or she would like to discuss the problem with a clinician or other service provider and to refer the case if so authorized (or if the person lacks the capacity to make an informed decision). Even if counseling is declined, the person should be given written information regarding available counseling and other services. Further protective action, including reporting to public agencies, should be undertaken only if the researcher (or the consulting clinician) has a substantial basis for believing that the elder person is in immediate danger of serious harm.

These responses require delicate judgment and clinical sensitivity. It is therefore in the principal investigators' interests to train their interviewers carefully and to specify procedures in advance for notifying the principal investigator of disturbing disclosures of past mistreatment and concerns about future danger. These concerns would be most likely to arise in faceto-face interviews. Information elicited in a telephone interview is highly unlikely to be so suggestive of future victimization as to warrant disclosure without consent.

A final question is whether the prospect of referral (without consent) just described (or whatever approach is prescribed in the protocol) should (or must) be disclosed to the prospective participants during the consent process. The proper response probably varies across research contexts. The Common Rule provides that an investigator seeking IRB approval to omit information that ordinarily must be disclosed must establish that participants will be exposed to no more than minimal risk; that waiving or altering informed consent “will not adversely affect the rights and welfare of the subjects;” and that the study would be “impracticable” if the information had to be disclosed. These criteria raise two questions:

First, does the residual possibility that the investigators or others could notify authorities, leading to unwanted intervention, family disruption, or other harm, amount to “more than a minimal risk” of harm to the participant? As Dresser (this volume) suggests, the answer to this question probably turns on whether the research is undertaken in a clinical or social service setting in which the risk (of disclosure, especially under applicable reporting laws) inheres in the clinical or service interaction, not in the research itself. In such instances, the residual possibility that the researcher might disclose evidence of mistreatment does not present a material risk that must be disclosed in order to obtain informed consent. As Dresser observes:

In applying the Common Rule's minimal risk and consent waiver provisions, the focus should be on whether research participation exposes older persons and caregivers to risks greater than those present in ordinary life and routine medical and social service encounters. For example, consider an interview study of family members caring for older individuals. If the study would expose participants to more detailed scrutiny than they would encounter in their usual interactions with the health care and social services systems, and if investigators planned to report suspected maltreatment to protective services authorities, the study would present more than minimal risk to study participants. The higher risk would exist because study participation would expose family members to reporting risks greater than those present in routine clinical and social services activities. If, however, data would be collected in interviews conducted as part of the ordinary activities of a social services agency, and the agency's ordinary reporting practices would be followed, the research risks would appear not to exceed the minimal risk threshold.

Assuming that the risk is characterized as a minimal one, the second question is whether disclosing it would unduly compromise the scientific integrity of the study by skewing enrollment or the accuracy of responses. Obviously, this question must be addressed in the context of a particular study. However, waiver of the disclosure requirement would seem to be especially warranted in studies aiming to provide accurate estimates of the occurrence of elder mistreatment.


In general, it is the panel's view that investigators and IRBs need clearer guidance (without rigid rules) concerning two issues that tend to recur in elder mistreatment research: conditions under which research can properly go forward with participants whose decisional capacity is impaired, and the proper responses to evidence of mistreatment elicited during the course of the study. In the absence of better guidance, IRBs are left setting their own criteria, leading to inconsistencies and confusion—the same IRB often interprets the governing rules differently with each rotation of its chair. Cooperative research between agencies or organizations is also difficult, if not impossible, since different IRBs often take different positions on these issues, including what information must be disclosed to obtain informed consent.

As a first step in this direction, the panel has sought to clarify some of the issues in these two areas and to provide some needed guidance. Eventually, the National Institute on Aging, in collaboration with the Office for Human Research Protections and other federal partners, should take the lead to promote further clarification, thereby helping investigators and IRBs to achieve the proper level of participant protection while enabling important research involving older and vulnerable adults to move forward. In addition, NIH should continue to support research on assessing decisional impairments and the adequacy of participant consent, on the effects of interventions aiming to facilitate and sustain consent by subjects with impairments, and on the nature and consequences of investigator and IRB responses to disclosures of mistreatment during the course of a study.



There are no cases involving elder reporting obligations. However, the attorney general of Iowa, in Opinion 83-11-3 (1983), concluded that a federal confidentiality certificate superseded Iowa's mandatory state child abuse reporting requirements. The supreme court of Minnesota subsequently ruled that the preemptive effect of the federal law authorizing Health and Human Services to issue confidentiality certificates for alcohol treatment research was itself superseded by a subsequently enacted federal statute requiring states to enact child abuse reporting statutes as a condition of receiving federal funds (State v. Andring, 342 N.W. 2d 128, 1984). It should be noted, however, that federal law does not require states to adopt mandatory reporting as a condition of receiving federal funds for elder mistreatment or adult protective services.