Ethical and Policy Issues in Research on Elder Abuse and Neglect
Research on elder abuse and neglect poses a multitude of challenges. Besides presenting methodological and practical difficulties, studying maltreatment of older people raises formidable ethical and policy problems. Two general features of the research account for these problems. First, the study population includes older persons with various mental, physical, and social vulnerabilities. Second, the research involves collecting information that could have negative legal, financial, and social consequences for the older persons and caregivers being studied.
The ethical and policy analysis is further complicated by an absence of regulatory guidance and ethical consensus regarding the appropriate procedures to govern research involving persons who lack the ability to decide about research participation. A similar lack of guidance exists regarding the conduct of research in nursing homes and other residential facilities. Finally, current policy and ethics fail to resolve many questions about the appropriate approach to research seeking legally and socially sensitive information.
In this paper, I address ethical and policy issues raised by research on elder abuse and neglect. I use as a framework for ethical analysis the Belmont Report, a document that identifies ethical principles and guidelines
for research involving human participants1 (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979). I use as a framework for policy analysis the Federal Policy for the Protection of Human Subjects (U.S. Department of Health and Human Services, 1991), also known as the Common Rule. My analysis reviews general issues relevant to research on elder abuse and neglect. Because states vary in their approaches to regulating disclosures of private information, professional reporting duties, nursing home operations, and other relevant topics, issues raised by individual research projects must be separately evaluated by local institutional review boards (IRBs) and attorneys.
The Belmont Report describes the characteristic features of research involving human participants and articulates three ethical principles that apply to such research. These principles are (1) respect for persons, (2) beneficence, and (3) justice. The principle of respect for persons underlies the requirement for informed consent to study participation. The principle of beneficence underlies the requirement to evaluate and balance risks and expected benefits in human studies. The principle of justice addresses fairness in selection of research participants. Provisions in the Common Rule incorporate these Belmont Report principles and requirements.
Institutional review boards rely heavily on the Belmont Report and Common Rule when they evaluate research proposals. Thus, studies on elder abuse and neglect must take into account the concepts and considerations in these documents. Although the Department of Health and Human Services has adopted additional regulations to cover certain populations deemed especially vulnerable in research, it has no special regulations governing research involving older persons, persons with impaired decision making capacity, or residents of nursing homes and other institutions. Various individuals and groups have, however, made recommendations addressing ethical issues with particular relevance to these populations.
In addition, the Common Rule contains a few provisions that specifically bear on vulnerable populations in research. The Rule directs IRBs to “be particularly cognizant of the special problems of research involving vulnerable populations” and to ensure that “[w]hen some or all of the subjects are likely to be vulnerable to coercion or undue influence, . . . additional safeguards have been included in the study to protect the rights and welfare of these subjects” (1991:28,016). The Common Rule advises IRBs that regularly review research involving a vulnerable subject population to consider including “one or more individuals who are knowledgeable about and experienced in working with these subjects” (1991:28,015).
In considering the issues raised by research involving elder abuse and neglect, one should keep in mind certain characteristics of U.S. policy governing research involving human subjects. Ethical principles and federal regulations establish a general framework for analyzing research proposals. In the current system, local IRBs, as well as funding agencies, interpret and apply the general principles and regulations to specific studies. The oversight system is based on the judgment that studies are sufficiently different and complex that it would be difficult (and probably futile) to set rules for every possible situation. Moreover, the current system reflects the government’s desire to avoid a centralized approach in which federal officials are responsible for reviewing study proposals. Instead, the system is designed to allow staff at the local institution and people from the local community to decide how the general principles should apply to individual studies. One inevitable consequence of this system is variation in IRB decisions addressing matters not definitively resolved by the federal regulations.
WHICH PROJECTS ARE GOVERNED BY RESEARCH ETHICS PRINCIPLES AND FEDERAL POLICY?
An initial task is to determine which projects qualify as research and which research projects are covered by the Common Rule. Not all information gathering and interventions related to elder abuse and neglect involve research. Moreover, some research is exempt from federal oversight. The Belmont Report and the Common Rule address, but do not fully resolve, these classification issues.
Distinguishing Research from Other Information Gathering Activities
Underlying the research oversight system is the judgment that research presents particular ethical concerns. Past incidents illustrate that the rights and interests of participants may be compromised in research. As a result, individuals are owed certain special protections in research that may not be required in other data-collection contexts. The Common Rule and its underlying principles are intended to cover the process of producing generalizable knowledge, an activity that society labels desirable but not so important that people should be compelled to participate (Pritchard, 2000). A project’s classification as research determines the nature of ethical and policy scrutiny it receives.
Data collection can occur in numerous contexts, including patient care, social services, public health, and program evaluation contexts. These activities fall under the general heading of practice and are not covered by the research oversight system. At the same time, these activities may be
combined with research. When they are, they should undergo the same ethical and policy assessment as other research projects.
The Belmont Report relies primarily on the different goals of research and practice to distinguish the two activities. Medical and behavioral clinicians gather information or perform interventions to advance the interests of individual patients or clients. Researchers, on the other hand, collect data and perform interventions “to test an hypothesis, permit conclusions to be drawn, and thereby to develop or contribute to generalizable knowledge” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:3). Like the Belmont Report, the Common Rule’s definition of research looks primarily to the intended aim of the activity. According to the Common Rule, “[r]esearch means a systematic investigation, including research development, testing, and evaluation, designed to develop or contribute to generalized knowledge” (U.S. Department of Health and Human Services, 1991:28,013).
The National Bioethics Advisory Commission (NBAC) recently called attention to flaws in the Common Rule definition. The NBAC observed that the definition omits the additional important point that research is done primarily to benefit society, while practice activities primarily aim to benefit patients, clients, or specific populations. In this respect, the NBAC noted, the Common Rule definition fails to recognize the possible conflict of interest that “always exists between investigators’ desires to pursue knowledge and their obligation to protect the rights and welfare of research participants” (2001:35). Moreover, the NBAC noted that the Common Rule definition provides little assistance to those seeking to distinguish research from activities such as public health and quality improvement projects.
As a result of these regulatory shortcomings, it can be difficult to determine when various data-gathering activities should be considered research. Three authors addressing this problem suggested that as a general rule, “[r]esearch projects are done to change the way the [health or social services] community thinks about a specific issue,” while “[n]onresearch investigations are done to give a specific group the information they need to make a specific decision” (Amdur et al., 2000).
Consistent with this approach, the Centers for Disease Control and Prevention Guidelines for Defining Public Health Research and Public Health Non-Research (1999) provide as follows:
The major difference between research and nonresearch lies in the primary intent of the activity. The primary intent of research is to generate or contribute to generalizable knowledge. The primary intent of nonresearch in public health is to prevent or control disease or injury and improve health, or to improve a public health program or service. Knowledge may be gained in any public health endeavor designed to prevent disease or
injury or improve a program or service. In some cases, that knowledge may be generalizable, but the primary intention of the endeavor is to benefit clients participating in a public health program or a population by controlling a health problem in the population from which the information is gathered (1999).
Similarly, a health care organization’s quality improvement efforts typically combine review of patient care data with a commitment to take corrective action if the findings warrant. When this commitment is absent, the organization’s information gathering is conducted primarily to benefit others and ought to receive the ethical scrutiny appropriate to a research project (Bellin and Dubler, 2001).
When public-health, quality-improvement, and other data-gathering projects have both research and practice objectives, they should be classified as research. Although individual participants or populations may directly benefit from their involvement in such projects, they also are subjected to such measures as structured interviews, intrusions on privacy, and potentially distressing questions to advance the aim of collecting generalizable data. When the aim of collecting generalizable data exposes individuals to risks or burdens not present in standard practice, an activity should be evaluated in the research oversight system (Cassarett et al., 2000). Because participants in such activities fail to receive services tailored to their individual needs and interests, and because the dominant goal is to produce knowledge for the benefit of others, these activities present risks and inconveniences to participants that are not typically present in the practice setting.
Determining Which Research Projects Are Subject to Federal Oversight
Determining whether a project involves research is just the first step in evaluating the project’s ethical and policy status. Not all activities that qualify as research are regulated by the federal government. The Common Rule explicitly applies only to research performed or funded by federal agencies that have adopted the Rule (U.S. Department of Health and Human Services, 1991:28,012). One of the Rule’s provisions seeks to extend this coverage, however, by requiring all institutions receiving federal research support to “protect the rights and welfare of human subjects of research conducted at or sponsored by the institution, regardless of whether the research is subject to federal regulation” (U.S. Department of Health and Human Services, 1991:28,014). In addition, some research conducted in the private sector is subject to explicit federal oversight. In separate regulations, the U.S. Food and Drug Administration (1999) requires manufacturers seeking agency approval for their products to conform their stud-
ies to provisions similar to those in the Common Rule. Certain states also have laws regulating the conduct of human studies (Glantz, 1992).
The Common Rule itself excludes certain research activities from its coverage. The Rule applies only to “research involving human subjects” (U.S. Department of Health and Human Services, 1991:28,012). According to the Rule, “human subject” is “a living individual about whom an investigator … conducting research obtains (1) data through intervention or interaction with the individual, or (2) identifiable private information” (U.S. Department of Health and Human Services, 1991:28,013). According to the Rule, information is private if it concerns “behavior that occurs in a context in which an individual can reasonably expect that no observation or recording is taking place” (U.S. Department of Health and Human Services, 1991:28,013). The Rule also classifies as private personal information that is collected for a particular purpose and that the individual reasonably believes will not be publicly disclosed, such as a medical record. When access to medical records or other private information would enable an investigator to readily determine the identity of a person, obtaining such information qualifies as research involving human subjects.
The Common Rule’s definition of research involving human subjects incorporates certain ethical judgments. First, if data are publicly available, researchers may use them without securing the permission of the individuals being studied (U.S. Department of Health and Human Services, 1991:28,012). For example, elder abuse and neglect research that involves the examination of public records, such as court or police records, is exempt from the Common Rule provisions. Second, if the information is not publicly available, researchers may gain access without a person’s permission as long as that person cannot be individually identified (U.S. Department of Health and Human Services, 1991:28,012).2 For example, elder abuse and neglect studies involving the examination of medical records lacking individual identifiers would be exempt from the Common Rule’s coverage. Another provision of the Common Rule exempts from its coverage surveys and interviews in which no identifying information is recorded (U.S. Department of Health and Human Services, 1991:28,012).
In adopting these provisions, federal officials determined that a person’s privacy interests do not extend to certain kinds of information (King, 1995). The Common Rule makes investigators and IRBs responsible for deciding whether people have reasonable expectations that personal information is
private and when access to private information would enable investigators to identify individuals.
RESPECT FOR PERSONS IN RESEARCH ON ELDER ABUSE AND NEGLECT
The Belmont Report’s principle of respect for persons expresses the moral judgment that no one should be used in research purely as a means to benefit others. In essence, the principle gives protection of individual rights and welfare priority over any medical or social benefits research might generate.
The Belmont Report describes two elements of the principle of respect for persons. The first element is that individuals capable of autonomous decision making should be permitted to make their own choices about whether to participate in research. To enable individuals to make autonomous choices, investigators must disclose important facts about a study, ensure that prospective participants understand that information, and ensure that decisions to enroll are not a response to undue pressures or incentives.
The second element of the principle is that individuals with impaired decision making capacities should be protected from harm in the research process. This dimension of the principle is reflected in legal requirements preventing investigators from enrolling decisionally incapable individuals in research without the informed consent of a family member or other appropriate person. Adequate protection may also require investigators to exclude decisionally incapable individuals from certain risky or burdensome studies that are permissible when conducted with capable, consenting individuals.
Standards and Procedures for Evaluating Decisional Capacity
Central to applying the principle of respect for persons is the determination of whether a prospective research participant can make autonomous decisions about study participation. Because capacity determinations have significant moral implications, evaluators must strive to avoid erroneous classifications. Treating a decisionally capable person as incapable leads to a demeaning and unjustified deprivation of that person’s right to decide whether research participation would be consistent with his or her particular values and preferences. Treating a decisionally incapable person as capable exposes that person to exploitation to advance the interests of those who benefit from the research enterprise.
In research on elder abuse and neglect, both types of errors can have serious consequences. Delegating research decision-making authority to
another individual can be especially problematic in this context, in light of the difficulties that may arise in selecting an appropriate research representative. Moreover, research on abuse and neglect can expose vulnerable individuals to physical, psychological, social, and other risks that ought not be assumed by someone who is unable to comprehend them or is acting in response to perceived pressure from others.
Although no single definition is enshrined in research ethics and policy, agreement exists on the basic features of decisional capacity. According to Paul Appelbaum and Thomas Grisso (1988), four abilities are relevant to decision-making capacity: (1) ability to communicate a choice, (2) ability to understand relevant information, (3) ability to appreciate how this information applies to one’s current and future situation, and (4) ability to give comprehensible reasons for a decision. Similarly, a Hastings Center group considering capacity to decide about life-sustaining treatment declared that a “patient has the capacity to make the treatment decision when he or she can understand the relevant information, reflect on it in accordance with his or her values, and communicate with caregivers” (1987:23).
The presence or absence of a dementia diagnosis fails to indicate whether someone has the necessary abilities to make choices about study participation. Similarly, although they may furnish preliminary guidance, simple mental status assessments are insufficient to determine whether someone is able to make autonomous research decisions. Instead, investigators should evaluate decisional capacity in the context of the specific study being proposed.
In the research setting, decision-making capacity should be assessed through a discussion of the facts relevant to the particular choice facing a prospective participant. At minimum, a prospective participant should demonstrate the ability to “understand the purpose, procedures, risks, benefits, and alternatives to participation in the study (including nonparticipation), express a choice about participation; and understand that refusal to participate involves no penalty or loss of benefits to which the person should otherwise be entitled” (Advisory Work Group on Human Subject Research, 1998:23). In this respect, assessing a person’s decision-making capacity goes hand in hand with assessing that person’s understanding of the important study information:
The concepts of capacity and adequate information are intertwined. To be informed, a subject must be cognitively capable of understanding the relevant facts about the decision at hand. To determine whether a subject has the requisite cognitive capacity, the examiner must disclose these facts and then ascertain the subject’s level of comprehension (Dresser, 1996:68).
As the NBAC pointed out, “our society has not decided what degree of impairment counts as a lack of decision making capacity” (1998:10). Many
groups support a sliding-scale approach to evaluating a person’s capacity to decide about research. According to this approach, capacity standards may be lower in research presenting comparatively low risk, while a higher level of decisional ability should be required for research that presents comparatively high risks, particularly when the research fails to offer participants the prospect of direct benefit (NBAC, 1998:24).
The assessment process itself can be conducted using a variety of methods. The simplest method is to ask prospective participants to describe briefly and in their own words the basic study information noted above (Wendler and Prasad, 2001). Richard Bonnie has suggested as options “specially tailored follow-up questions to assess subject understanding, videotaping or audiotaping of consent interviews, second opinions, use of consent specialists, or concurrent consent by a family member” (1997:110). The device of concurrent consent is often adopted in dementia research when prospective participants have uncertain or fluctuating decisional capacity (High et al., 1994). This will not always be an option in studies on elder abuse and neglect, but it could be feasible in some such studies. Another suggestion is to adopt a two-part consent process, in which relevant study information is presented and the prospective participant is then asked questions about the study. If the individual exhibits a lack of adequate comprehension, the information is presented again, and the individual is reexamined (Ratzan, 1985). Such a process can detect individuals with problematic memory or other incapacitating cognitive deficits.
The capacity examination should enable researchers to ascertain when someone lacks the requisite abilities to make an autonomous choice about study participation. For people who appear to have adequate decision-making capacity, the next step is to ensure that they actually understand the significant information relevant to becoming a study participant.
Informed Choices About Research
People deciding to enter a study should understand certain facts. The Belmont Report lists the following general items essential for investigators to describe: “the research procedure[s], their purposes, risks, and anticipated benefits, alternative procedures (where therapy is involved), and a statement offering the subject the opportunity to ask questions and to withdraw at any time from the research” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:5). The Report describes the following risks, each of which is specifically relevant to elder abuse and neglect research: “psychological harm, physical harm, legal harm, social harm, and economic harm” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:7).
The Common Rule includes a more detailed list of the information that prospective participants should understand. Two items are particularly relevant to research on elder abuse and neglect. First, researchers must describe “the extent, if any, to which confidentiality of records identifying the subject will be maintained” (U.S. Department of Health and Human Services, 1991:28,016). Second, if the research presents more than minimal risk, prospective subjects must be told whether compensation or treatment, or both, will be available if injury occurs (U.S. Department of Health and Human Services, 1991:28,016).
Helping prospective participants to achieve adequate understanding is no small task. Surveys and interviews indicate that an appreciable number of people fail to understand important information about the studies in which they participate. A major problem is that participants often mistakenly think that studies are done primarily for their individual benefit, rather than to advance knowledge (Kass and Sugarman, 1996). The challenges may be especially great in research on elder abuse and neglect, given the educational levels and health problems of many prospective participants. Researchers could also encounter difficulties in ensuring that participants understand when elder abuse and neglect studies offer them no direct personal benefit.
Besides the methods described previously in the discussion of capacity assessment, investigators may enhance participants’ comprehension by using graphics, videotapes, and other creative approaches to information disclosure (Sachs and Cassel, 1990). The NBAC suggested additional measures to enhance prospective participants’ understanding. These include the use of translators when investigators and participants speak different languages, seeking advice from representatives of study populations regarding the appropriate content and presentation of information to prospective participants, and focusing less on the consent form and more on ensuring an effective disclosure and decision-making process (2001:88, 100).
Voluntary Choices About Research
According to the Belmont Report, decisions to enroll in research are voluntary if they occur in the absence of coercion or undue influence. Coercion exists “when an overt threat of harm is intentionally presented by one person to another in order to obtain compliance.” Undue influence exists when there is “an offer of an excessive, unwarranted, inappropriate, or improper reward or other overture in order to obtain compliance” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:6).
A variety of factors may compromise the voluntariness of decisions to participate in elder abuse and neglect research. Older people who depend
on health and social services professionals for assistance may agree to participate out of a desire to preserve good relationships. Older people may fear they will lose needed services if they refuse to participate. Residents of nursing homes may be especially vulnerable to these pressures. Offers of monetary or other incentives for study participation may also be unduly tempting to economically disadvantaged people.
The NBAC suggested ways to address some of these problems. This group’s suggestions were to discuss possible research participation out of the presence of those to whom the prospective participant usually defers and to design studies so that staff in residential facilities are unaware which individuals are participating (2001:89). Others suggest that research discussions with nursing home residents be conducted in private and that the residents’ own physician and nurses not be involved in conducting the research (Sachs et al., 1993). If a physician is conducting research, the Helsinki Declaration urges particular caution and advises that in such a case, “the informed consent should be obtained by a well-informed physician who is not engaged in the investigation and who is completely independent of this relationship” (World Medical Association, 2000:3044).
Procedures for Assessing the Quality of Research Decisions
Investigators must adopt procedures for determining whether research choices are sufficiently capable, informed, and voluntary. At minimum, investigators should develop an assessment plan and the IRB should review that plan. Some individuals and advisory groups believe the assessment should be performed by a qualified professional not otherwise involved in the research. This proposal responds to the concern that researchers eager to enlist participants may be insufficiently demanding about the level of capacity, understanding, and voluntariness necessary to consent.
Some groups say that IRBs should be authorized to decide when independent evaluations are needed in specific studies. The NBAC declared that when research proposals present greater than minimal risk, IRBs ordinarily “should require that an independent, qualified professional assess the potential subject’s capacity to consent” (1998:58). In such cases, the proposal “should describe who will conduct the assessment and the nature of the assessment” (1998:58).
Although the use of independent evaluators will add to the costs of a study, supporters contend that the costs are justified by the need to protect individual rights and welfare. This is particularly true, they say, when individuals are asked to join studies that will expose them to significant risk (Dresser, 2001a).
Decisionally Incapable Persons
The screening process may identify individuals lacking decisional capacity. According to the principle of respect for persons, individuals incapable of autonomous choice should be protected from harm in the research process. Conferring such protection may require excluding them from research altogether, or excluding them from studies that present relatively high risk, approaches that are discussed below. If decisionally incapable individuals are considered for research participation, special protective measures are warranted. The most common protective safeguards are (1) to designate as a research decision maker someone “likely to understand the incompetent subject’s situation and to act in that person’s best interest” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979), and (2) to require the decisionally incapable person’s assent or lack of objection to study participation.
Selection of a Research Representative
The Common Rule allows a decisionally incapable person’s “legally authorized representative” to consent to that individual’s research participation, as long as the representative’s consent is adequately informed and voluntary. The Rule fails to specify who may assume the position of representative; instead, it refers generally to “an individual or judicial or other body authorized under applicable law to consent on behalf of a prospective subject to the subject’s participation in the procedure[s] involved in the research” (U.S. Department of Health and Human Services, 1991:28,013).
State laws provide for the appointment of legal guardians to make personal and financial decisions for individuals a court has declared legally incompetent. Some states also have laws explicitly authorizing courts, guardians, or family members to act as research decision makers, but many states lack clear rules in this area (Hoffman and Schwartz, 1998). In states without clear rules, possible decision makers include a court-appointed guardian, an individual the currently incapable person previously chose as a research or health care decision maker (also known as a research or health care proxy decision maker), or a relative or close friend of the decisionally incapable person (also known as a surrogate decision maker).3
In practice, people from all three of these groups make research choices
for decisionally incapable individuals. Indeed, relatives acting as informal surrogates are probably the most common research decision makers for older people unable to make their own choices. Limiting research decision making to court-appointed guardians is generally seen as burdensome, expensive, and unnecessary to supply adequate protection. Because guardianship proceedings typically fail to examine a prospective guardian’s qualifications to act as a research decision maker, such proceedings are an ineffective research safeguard. Moreover, court-appointed guardians may be strangers lacking knowledge of an incapable person’s current situation and former values and preferences.
At the same time, the practice of relying on informal surrogates for research decision making raises its own concerns. The primary worry is that such surrogates may have personal interests that conflict with those of the prospective research subject (Dresser, 1996). Conflicts of interest are an obvious possibility in research on elder abuse and neglect, particularly when an abusive or neglectful relative acts as the surrogate decision maker for an older incapable participant.
Investigators should make every effort to select research representatives genuinely concerned with the decisionally incapable individual’s welfare. One way to do this is to conduct screening and education of potential representatives, with the aim of detecting inappropriate decision makers and increasing the odds that those chosen will adequately protect research participants’ interests. Another is to ensure that the incapable person’s research representative will not be the person asked to supply information relevant to abuse and neglect.
The U.S. Department of Health and Human Services regulations on research involving children suggest a third possible response to the conflict of interest problem. The regulations provide as follows:
if the IRB determines that a research protocol is designed for conditions or for a subject population for which parental or guardian permission is not a reasonable requirement to protect the subjects (for example, neglected or abused children), it may waive the [usual parental] consent requirements . . . provided an appropriate mechanism for protecting the children who will participate as subjects in the research is substituted, and . . . that the waiver is not inconsistent with federal, state, or local law (1999, §46.408(c)).
The regulations also state that the “choice of an appropriate mechanism would depend upon the nature and purpose of the activities described in the protocol, the risk and anticipated benefit to the research subjects, and their age, maturity, status, and condition” (1999, §46.408(c)).
In discussing this research situation, the National Commission for the
Protection of Human Subjects of Biomedical and Behavioral Research suggested the following option:
to appoint a social worker, pediatric nurse, or physician to act as surrogate parent when the research is designed, for example, to study neglected or battered children. Such surrogate parents would be expected to participate not only in the process of soliciting the children’s cooperation but also in the conduct of the research, in order to provide reassurance for the subject and to intervene or support their desire to withdraw if participation becomes too stressful (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1975:19).
An analogous approach might be adopted in studies of elder abuse and neglect involving decisionally incapable adults. Major considerations would be whether it would be justifiable or practical to withhold information about the appointment of such a designated surrogate from a close family member suspected of abuse. As two authors addressing the child abuse research regulation noted, “removing the parent’s decision-making authority will not assuage a child’s concerns about the consequences of disclosure, or prevent a parent from vigorously asking a child about what was discussed in interviews . . .” (King and Churchill, 2000:722). Similar concerns could arise in research involving older adults with impaired decisional capacity. Other important considerations would be the study’s potential risks and direct benefits, variables that are addressed in the section on beneficence in research on elder abuse and neglect.4
Assent and Absence of Objection
Decisionally incapable persons ought not be completely excluded from research decision making. According to the Belmont Report, the principle of respect for persons “requires giving [such persons] the opportunity to choose to the extent they are able, whether or not to participate in re-
search” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:6). This requirement has two components. One is to supply information and help decisionally incapable persons understand as much as they can about the proposed study experience. The other is to respect their verbal or physical opposition to participation.
The assent requirement is met when a decisionally incapable person verbally agrees to cooperate with study procedures. The lack of objection requirement is met when a decisionally incapable person fails to indicate verbal or physical opposition to study participation. Assent or lack of objection alone is insufficient to authorize an incapable person’s research participation; the informed and voluntary permission of the legally authorized representative is required as well. The presence of the decisionally incapable person’s assent or lack of objection is an additional precondition to proceeding with research.
The demand for assent or lack of objection rests on the moral judgment that vulnerable incapable individuals ought not be forced to contribute to the generation of knowledge for the benefit of others. Instead, imposed interventions are warranted solely when they are absolutely necessary to protect the objecting individual’s personal welfare.
These moral judgments account for the current view that assent or lack of objection are mandatory in nearly all studies seeking the involvement of decisionally incapable persons. The one possible exception arises when research offers participants a direct health or behavioral benefit unavailable outside the research context. The rationale is that in this situation, “the decision is more akin to a treatment decision, and thus incompetent refusals may be overridden in some cases” (Berg, 1996:24). Although some individuals and groups support such an exception, others question its justification. As one group noted, “it is difficult to equate an intervention which is investigative in nature—with an intervention ‘which would be ordered in a purely therapeutic context’” (Keyserlingk et al., 1995:342). Compelling direct benefit would be necessary to justify including actively resisting incapable individuals in research on elder abuse and neglect.
Exceptions to the Requirement for Informed Consent to Research
The Belmont Report acknowledges that imposing the customary informed consent requirements could eliminate or seriously impede some research projects. The Report suggests that incomplete information disclosure could be acceptable under certain circumstances. This view is incorporated in a Common Rule provision permitting exceptions to the general requirements for informed consent. An investigator seeking IRB approval to omit or alter information that must ordinarily be disclosed, or to waive
the informed consent requirement altogether, must establish that (1) participants will be exposed to no more than minimal risk, (2) waiving or altering informed consent “will not adversely affect the rights and welfare of the subjects,” (3) the study would be impracticable if the information had to be disclosed, and (4) in appropriate cases, the withheld information will be disclosed at the end of the study (U.S. Department of Health and Human Services, 1991:28,016).
According to the Belmont Report, withholding information is never justified simply to make it easier to enroll participants. Instead, withholding should be permitted only if full disclosure would “destroy or invalidate the research” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:6). This suggests that IRBs should be rigorous in demanding investigators show that their studies would be impracticable if the usual informed consent requirements were applied.
An investigator proposing a study on elder abuse and neglect might seek to withhold information about the study, such as its purpose or risks, from prospective participants or research representatives. For example, an investigator might propose to omit from the consent discussion and form a statement that evidence of possible abuse or neglect will trigger a referral to the appropriate adult protective services agency. The investigator might argue that disclosure of this information would generate too many refusals to participate or distort the responses of those who agree to participate. Another example would be an investigator who sought to modify the study description so that prospective participants would be unaware that its focus was elder abuse or neglect.
Withholding study information from prospective participants compromises their right to make informed research decisions. Thus, it should be done only for compelling reasons. One relevant IRB consideration in the above examples would be the degree of risk that the proposed study would create for participants. If study procedures would expose participants to more than minimal risk, the IRB ought not approve the information withholding. (This topic is discussed in greater detail in the section on balancing harms and benefits in research, below.) The IRB would also need to evaluate the investigator’s claim that full disclosure would make the research impracticable.
Surprisingly, data from a study of parents’ and teenagers’ preferences suggest that disclosure of reporting plans might actually be a positive factor in participant recruitment. O’Sullivan and Fisher found that parents and adolescents would be more willing to enroll in studies on abuse and related behaviors if the investigator planned to notify the parents or help the child obtain independent assistance for any problems discovered in the course of the study (O’Sullivan and Fisher, 1997). These data suggest that IRBs should require persuasive evidence that research will be unduly compro-
mised before granting requests to withhold information from prospective participants in elder abuse and neglect studies.
Choices to participate in research on elder abuse and neglect should be capable, informed, and voluntary. When someone is unable to make valid decisions to participate, researchers should obtain the capable, informed, and voluntary permission of a relative or other individual concerned with protecting the incapable person’s well-being. Researchers should also seek the willing cooperation of the decisionally incapable person. Departures from these rules require compelling justification, including a showing that research interventions will not expose participants to significant risk.
BENEFICENCE IN RESEARCH ON ELDER ABUSE AND NEGLECT
The Belmont Report describes two dimensions of the beneficence principle in research. One is to avoid harm to study participants, and the other is to maximize possible benefits and minimize possible harms. Risks to research participants cannot be completely avoided. Instead, those evaluating the ethics of proposed studies must “decide when it is justifiable to seek certain benefits despite the risks involved, and when the benefits should be forgone because of the risks” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:4)
Two Common Rule provisions incorporate the beneficence principle. The first provision directs researchers to minimize risks to participants through the use of procedures that are “consistent with sound research design,” do not impose “unnecessary risks,” and, when possible, are “already being performed on the subject for diagnostic or treatment purposes” (1991:28,015). The second provision requires researchers to justify any unavoidable risks by discussing the benefits a study is expected to produce. According to the Common Rule, risks must be “reasonable in relation to anticipated benefits, if any, to subjects, and the importance of the knowledge that may reasonably be expected to result” (U.S. Department of Health and Human Services, 1991:28,015).
To apply the beneficence principle in research, investigators and IRBs must first determine the risks and potential benefits presented by specific studies. The Belmont Report directs study reviewers to be systematic in evaluating potential harms and benefits:
The method of ascertaining risks should be explicit, especially where there is no alternative to the use of such vague categories as small or slight risk. It should also be determined whether an investigator’s estimates of the probability of harm or benefits are reasonable, as judged by known facts
or other available studies (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:7).
The material below describes types of harms and benefits possible in elder abuse and neglect research.
Potential Benefits Offered by Research
Research benefits are available primarily to two groups. First, people enrolled in elder abuse and neglect studies might receive personal benefits due to their study participation. Second, as noted earlier, research is conducted primarily to generate knowledge that could contribute to improved medical and social interventions for others.
Benefits to research participants are generally classified as direct or indirect. Biomedical and behavioral studies may offer a variety of direct benefits to participants. As the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research noted, “[t]o be considered ‘direct,’ the possibility of benefit to the subject must be fairly immediate [and t]he expectation of success should be well-founded scientifically” (1978:13).
Examples of research offering potential direct benefit are studies designed to enhance current biomedical or behavioral therapies or to create new training or educational materials (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1978:13). Direct benefits in research on elder abuse and neglect could be available in studies comparing different interventions to reduce the occurrence of elder abuse and neglect or studies to determine the best methods of assistance to provide once maltreatment has occurred.
To be considered direct, potential benefits should be provided by the intervention(s) under study. In contrast, indirect benefits may be received independently of the primary study interventions (King, 2000). Such benefits may include added contact with health or social services professionals, opportunities to engage in social activities outside of one’s usual routine, and a sense of satisfaction in contributing to a valuable social endeavor. Indirect benefits can be difficult to predict due to variations in individual responses. For example, although some older people may value the above elements of the research experience, others may be distressed by having their routines disrupted or receiving additional attention from health or social services professionals. For this reason, although indirect benefits may “count” as possible benefits to study participants, they ought not be assigned the same significance as potential direct benefits when studies are discussed with prospective participants or reviewed by IRBs (Keyserlingk et al., 1995). The NBAC asserted that indirect benefits should be recognized,
but “should not weigh in the judgment of IRBs regarding the balance of risks and potential benefits to the participants” (2001:74). A similar approach should be taken regarding benefits studies offer to participants’ relatives or other caregivers.
It is generally agreed that items such as financial incentives and reimbursement for transportation and other costs associated with research participation should not be considered benefits to participants. If payments were allowed to count as benefits, then a high payment could offset serious and otherwise unacceptable risks to participants (NBAC, 2001:74).
The evaluation of a study’s potential benefits should also consider its possible benefits to society. Evaluating possible societal benefits involves scrutinizing study design, personnel, and other factors affecting the quality of information collected. Because poorly designed or conducted studies will not yield valid and reliable data, they fail to offer benefits to others. Involving people in low-quality studies imposes burdens and risks on them without an adequate social justification. Thus, investigators must propose, and IRBs must demand, studies that meet the relevant scientific standards. When studies are not expected to undergo rigorous peer review as part of the funding process, investigators and IRBs have a responsibility to obtain such a review (Office of Protection from Research Risks, 1993:4–11).
Another dimension of research benefit to others concerns the importance of the research aim. This dimension of research value has received relatively little attention in research ethics analysis, but it deserves more serious consideration. In elder abuse and neglect studies, as in all human research, there must be a good reason for exposing research participants to inconveniences, burdens, and possible harms. Investigators should be able to show why the specific study problem is socially significant and how information gained from the study will help address the problem (NBAC, 2001:73). For example, studies proposing to test minor variations in existing health or social service interventions might hold little prospect of material benefit to others.5
Potential Harms in Research
As noted above, the Belmont Report lists the following as risks to research participants: “psychological harm, physical harm, legal harm, social harm, and economic harm” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:7). More recently, the NBAC elaborated on these concepts. Included as physical
harms are “injury, illness, pain, suffering, or discomfort.” Types of psychological harm are “negative perceptions of self, emotional suffering (e.g., anxiety or shame), or aberrations in thoughts or behavior.” The Commission defined social harms as “negative effects on one’s interactions or relationships with others.” Participants experience economic harm when they incur expenses as a direct or indirect result of research participation. Legal harms arise when research participation exposes people to actions such as “arrest, conviction, incarceration, or lawsuits” (NBAC, 2001:71–72).
These categories of research risk could exist in studies of elder abuse and neglect. Older participants in such studies could suffer physical harm from family or professional caregivers due to the participant’s negative statements about these individuals. Older participants could also experience anxiety or other emotional distress at being asked about a difficult living situation. Relationships with relatives could be damaged as a result of information provided to researchers. Older participants could incur financial costs if the research triggers removal of an unpaid family caregiver or another change in living situation. Information disclosures to researchers could also lead to guardianship or other legal actions, which could be perceived as detrimental by older participants.
Family and professional caregivers participating in elder abuse and neglect studies could face an array of research harms as well. Although they would be unlikely to experience physical harm as a result of research, shame and other psychological harm from being asked to discuss potentially problematic behavior could be a common risk in such studies. In addition, family relationships could suffer if research triggers questions about a caregiver’s conduct or abilities. Family caregivers could be deprived of financial benefits if research leads to removal of relatives from an older person’s home. Professional caregivers could experience economic harm if their research participation leads to employment termination. Persons reported for abuse and neglect based on information elicited in research could also be subjected to legal harm.
Of course, older persons and their caregivers face risks of this sort in their customary interactions with health care and social services personnel. In evaluating potential research harms, the emphasis should be on any additional risks that will arise due to the data gathering associated with a research project. If research participation will entail interviews, examinations, or other interventions that participants would not encounter in their usual contacts with health care and social services personnel, the risks accompanying these interventions are properly classified as research risks that should be minimized, justified, and, in many cases, disclosed to potential participants.
Minimizing Risk in Research
Minimizing physical and other research-related risks to participants in elder abuse and neglect studies can raise complex issues. The Common Rule directs researchers to minimize risks “by using procedures already being performed on the subjects for diagnostic or treatment purposes” (U.S. Department of Health and Human Services, 1991:28,015). Investigators may also reduce risks by ensuring that (1) the research team is properly qualified and trained, (2) a system is in place to address harm arising in the research process, (3) confidentiality measures are appropriate, (4) participants are adequately monitored during the study, and (5) individuals are excluded or withdrawn from studies presenting unacceptable risk (NBAC, 2001:73).
Two general features of elder abuse and neglect research pose risks to participants. First, investigators often seek information that could produce harmful consequences if it were released to protective services or other authorities. A variety of methods allow investigators to obtain needed data while minimizing this risk. Sometimes data can be obtained from public records or other public sources. Sometimes data can be collected through the use of interviews or other procedures conducted in routine health care and social services activities; in such cases, data collection will not expose individuals to reporting risks beyond those they would encounter in clinical and social services settings. Sometimes data can be recorded without identifying participants. Sometimes identifying information can be destroyed shortly after it is collected. Strict security measures can also be adopted, including identifying participants by code and severely restricting access to documents linking codes with individual participants (Office of Protection from Research Risks, 1993:3–33). The Common Rule permits IRBs to waive the usual requirement for a signed consent form if such forms would be the sole record tying an individual participant to research and a breach of confidentiality constitutes the principal study risk (U.S. Department of Health and Human Services, 1991:28,017).
When investigators studying elder abuse and neglect seek information about a participant’s mental health, illegal conduct, drug or alcohol use, or other conduct with potential negative financial or employment consequences, they may obtain special confidentiality protections from federal officials. Federal law authorizes the Secretary of the Department of Health and Human Services to issue certificates of confidentiality to protect participants’ privacy in both government and privately funded projects (Public Health Service Act §310(d), U.S. Code 42 §241(d), 1994). The certificates allow investigators to withhold “from all persons not connected with the conduct of such research the names or other identifying characteristics of such individuals.” Investigators with certificates “may not be compelled in
any federal, state or local civil, criminal, administrative, legislative, or other proceedings to identify” research participants (Public Health Service Act, 1994). Certificates are granted only when officials deem them necessary to achieve the project aims.
Certificates of confidentiality may prevent researchers from being forced by government authorities to disclose private information about participants suspected to be victims or perpetrators of elder abuse and neglect. Yet some investigators and commentators question whether certificates legally exempt researchers from reporting requirements, and no definitive court ruling exists to resolve this question (Amaya-Jackson et al., 2000). Some federal agencies reportedly refuse funding for studies in which investigators with confidentiality certificates plan to refrain from reporting child abuse and neglect (Runyan, 2000). Moreover, confidentiality certificates do not prohibit investigators from voluntarily disclosing information about suspected abuse and neglect. Thus, researchers must address the ethical issues that arise when withholding information could leave older persons in serious jeopardy.
This point is related to the second risk-producing feature of elder abuse and neglect research. Older persons who participate and those whose relatives or professional caregivers participate may face physical and other serious risks. To minimize such risks to older persons, researchers should take measures to address suspected abuse and neglect. But such measures may cause harm to relatives or professional caregivers. Moreover, such measures may be opposed by older participants who prefer their risky situations to the available alternatives. Finally, social services interventions may do more harm than good for an older individual.
As this discussion reveals, the ethical and policy directive to minimize risks can present difficult trade-offs in elder abuse and neglect research. Measures to minimize harm to older participants, such as reporting suspected abuse and neglect to authorities and the protective interventions that follow, can increase psychological, social, financial, and legal risks to relatives and caregivers. Similarly, measures to safeguard private information about families and caregivers can leave older persons vulnerable to harm. And measures to minimize certain kinds of harm to older persons may increase their exposure to other kinds of harm. The conflicts that may arise in attempts to protect older persons and caregivers from research harms are addressed in further detail in the section on confronting major ethical conflicts.
Balancing Potential Harms and Benefits in Research
After investigators have reduced study risks as much as possible, some risks will almost always remain. The IRB must then determine that these remaining risks are justified by the benefits the study is expected to yield.
As noted above, the Common Rule requires that a study’s risks are “reasonable in relation to anticipated benefits.” “Reasonable” is a term subject to varied interpretation. In directing IRBs to decide whether a proposed study presents a reasonable balance of risks and anticipated benefits, federal officials delegated to IRB members the authority to interpret what risks are reasonable. The Common Rule requires that IRBs include as members nonscientists and persons not affiliated with the research institution (U.S. Department of Health and Human Services, 1991:28,015). By requiring such members, officials sought to ensure that reasonableness will be evaluated not just by researchers and their colleagues, but by persons who might have a different perspective on research harms and benefits. Reform proposals to increase the number of nonscientist and unaffiliated members are intended to increase the chance that a diversity of perspectives will be represented when research harms and benefits are balanced (NBAC, 2001:63).
Besides requiring IRBs to decide whether research risks are reasonable, the Common Rule and the federal regulations addressing research involving vulnerable populations direct IRBs to determine when research risks are minimal. According to the Common Rule, study interventions present minimal risk if “the probability and magnitude of harm or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests” (U.S. Department of Health and Human Services, 1991:28,013–28,014).
As observed in the section on respect for persons in research, the Common Rule permits investigators to withhold certain information from prospective participants, or to forgo consent altogether, only if research interventions present no more than minimal risk. In addition, the NBAC proposed that research interventions presenting greater than minimal risk and no prospect of direct benefit should be allowed only with an incapable individual’s prior competent consent or after review and approval by a special standing panel of diverse experts and community representatives convened by the Secretary of the Department of Health and Human Services (NBAC, 1998:54, 61).
The above approaches require IRBs and other reviewers to apply the Common Rule’s minimal risk definition. Not everyone is satisfied with the regulatory definition, however. Some argue that its reference to the risks of everyday life is problematic. These critics say that IRB members are unlikely to know enough about the nature, probability, and degree of everyday risks to make comparisons with research risks (Kopelman, 1995). Furthermore, empirical studies have shown variations in the research interventions clinicians and others classify as minimal risk (Dresser, 1999:22).
Another problem is that the Common Rule’s definition of minimal risk fails to clarify whether research risks should be compared to everyday risks
faced by people in the general population or by subjects individually. Federal regulations issued in 1981 were accompanied by a preamble stating that “the risks of harm ordinarily encountered in daily life means those risks encountered in the daily lives of the subjects of the research” (U.S. Department of Health and Human Services, 1981:8373). But this interpretation is contested by the NBAC (2001:83) and other advisory groups (National Human Research Protections Advisory Committee, 2001), which take the position that minimal risk should be tied to daily risks faced by members of the general population. If everyday risks were individualized, higher research risks could be labeled minimal risk for participants facing higher-than-average daily risks due to illness or poor living situations. As a result, less ethical scrutiny could be devoted to studies involving people already at relative disadvantage. For example, a waiver of informed consent might be obtained in a study presenting higher risk because it involved participants living in high-crime areas. And as the NBAC noted, the individualized standard for minimal risk “would impose disproportionate burdens of research on the ill and provide weaker protections for them than for healthy individuals” (2001:83). Such an approach would thus disregard the ethical principle of justice, discussed below.
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 services 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. In the first situation, the Common Rule would appear to rule out a consent waiver; in the second, the minimal risk requirement for a waiver would be satisfied.
An appropriate balance of risks and expected benefits is central to the ethical conduct of research on elder abuse and neglect. Specific challenges in determining what constitutes such a balance are discussed below in the section on confronting major ethical conflicts.
JUSTICE IN RESEARCH ON ELDER ABUSE AND NEGLECT
The Belmont Report’s final ethical principle is justice. Justice in research involves the fair distribution of risks and benefits associated with research. The principle advises against relying too heavily on disadvantaged individuals or groups to bear the burdens of acting as research participants. On the other hand, the justice principle holds that individuals and groups should receive a fair share of the benefits available through study participation and the improved health and social interventions research makes possible.
The justice principle has several applications to research on elder abuse and neglect. One is that investigators observing the principle will not seek study participants solely from economically and socially disadvantaged populations simply because of “their easy availability, their compromised positions, or their manipulability” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:5). Similarly, vulnerable populations such as decisionally incapable adults should not be involved in research if adequate data can be obtained by studying adults able to make their own choices about participation. Instead, there must be sound scientific and policy justifications for targeting disadvantaged or vulnerable populations in research.
Three additional applications of the justice principle pertain to the distribution of benefits available through research on elder abuse and neglect. One such application is to give individuals and groups fair access to studies that offer potential direct benefit to participants. Thus, investigators ought not exclude certain people from studies simply because it is more inconvenient or costly to recruit them (Institute of Medicine, 1994:82).
Another application of the justice principle involves ensuring that the improved interventions publicly funded research makes possible are reasonably available to research participants and to those in the general population who would benefit from them (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979:5). Thus, when elder abuse and neglect studies yield health and social benefits, those benefits should be available to groups that participated in the studies and to others based on need rather than ability to pay.
Finally, the failure to implement a robust research program to study elder abuse and neglect can be seen as unjust, because it denies vulnerable older persons and their families the benefits that could flow from such a program. A fair system of research funding will allocate funds to studies of serious public health needs, including elder abuse and neglect (Dresser, 2001b).
CONFRONTING THE MAJOR ETHICAL CONFLICTS IN RESEARCH ON ELDER ABUSE AND NEGLECT
The societal importance of advancing knowledge and developing improved health and social programs addressing elder abuse and neglect is undeniable. At the same time, other important interests must be respected in the research process. Research ethics and policy recognize the interests of individuals (or their authorized representatives) in being informed about the studies they are asked to join. Research ethics and policy also deem important the protection of research participants from unnecessary and unreasonable risks. These ethical and policy considerations support certain restrictions on the conduct of research. Such restrictions reflect the judgment that advances in knowledge fail to justify compromising significant interests of the individual.
In research on elder abuse and neglect, the most serious ethical conflicts involve societal interests in gaining knowledge about elder abuse and neglect, participants’ interests in being informed of research risks, and the interests of both participants and nonparticipants in being protected from harm. Below, I discuss the major conflicts that can arise and offer guidance on how to resolve them.
Conflicts Regarding Responses to Suspected Abuse and Neglect
The first set of conflicts concerns researchers’ responses to suspected elder abuse and neglect. Here, a variety of interests may be at odds. Societal interests in advancing knowledge might be best served if researchers take no action in response to suspected harmful conduct. Relatives, professional caregivers, and older persons who prefer their current living situations might be more likely to enroll and supply accurate information if the information will be used purely for research purposes.
Though this approach promotes societal interests in generating high-quality research data, it gives little regard to older persons’ interests in protection from harm. From a broader perspective, the approach overlooks the societal interest in preventing elder abuse and neglect, an interest explicitly recognized in laws requiring reporting of suspected maltreatment.
A researcher’s complete failure to respond to suspected abuse and neglect is questionable on both legal and ethical grounds. Although state reporting laws vary, some have language that could apply to researchers (Garfield, 1991). As noted above, a federal certificate of confidentiality does not necessarily exempt researchers from state reporting requirements. Apart from statutory reporting mandates, some writers suggest that courts could hold investigators legally responsible for acting to minimize harm they discover in the course of research. The basis of this duty would be the
decision in Tarasoff v Board of Regents of the University of California (529 P2d 334, 1976) and subsequent court decisions holding that professionals have duties to take reasonable measures to protect potential victims of their patients’ or clients’ violence. Paul Appelbaum and Alan Rosenbaum note that courts might deem certificates of confidentiality irrelevant to the investigators’ duty to take some action on behalf of potential victims (1989).
Besides their possible legal obligations, researchers may have ethical responsibilities to respond to suspected neglect and abuse. Appelbaum and Rosenbaum suggest that “[o]n ethical grounds alone, . . . it may be reasonable for investigators to consider building provisions for the protection of potential victims into their protocols” (1989:889). Investigators in a consortium of longitudinal studies of child abuse and neglect known as LONGSCAN, as well as those conducting the MacArthur Risk Assessment Study of community violence in former mental patients, recognized an ethical responsibility to prevent harm to both study participants and third parties, although their exact responses varied due to the specific study circumstances and different philosophies of study teams and their reviewing IRBs (Knight et al., 2000; Monahan et al., 1993).6
The ethical and legal duties to report suspected abuse and neglect are strongest when researchers conclude that an older person is at risk of serious, imminent harm. If an older person appears to be in imminent danger, a report to protective services workers may be justified even if the older person refuses assistance (Garfield, 1991:870). Similarly, when investigators believe a decisionally impaired older adult is at risk of grave and immediate harm, they may have both legal and ethical responsibilities to report, despite a caregiver’s refusal of assistance. As the probability, magnitude, and imminence of harm decrease, however, so does the justification
for reporting. In the absence of a clear and serious threat, investigators may adopt alternative responses to the living situations of older persons and their caregivers.
One option is to furnish research participants with information on available community resources for individuals and families with various living difficulties. Investigators can also express any concerns they may have about a participant’s living situation and urge the at-risk individual to obtain help. In planning their studies, investigators can identify an appropriate services agency and establish a referral process for participants expressing a wish for assistance (Appelbaum and Rosenbaum, 1989:892).
Investigators in the LONGSCAN studies adopted a variety of responses to participants in research on child abuse and neglect. In these studies, research team members are most likely to learn of possible maltreatment through interviews with 12-year-olds. Before their study participation, parents and children are told that children will be asked about abuse, neglect, or exposure to violence and that investigators will report children deemed to be at risk of serious harm. Children participate in a computer-administered interview that includes questions about various forms of abuse. A human interviewer is present but is unable to see the child’s answers. The children are told that they may skip questions and stop the session at any time.
The LONGSCAN interviewers are trained to recognize and comfort children in distress, and interviewers may terminate the session when a child exhibits persistent distress. Interviews end with “a debriefing with each child that acknowledges that some of the questions may have been difficult, and all study participants are provided with a list of local family services agencies and how and why to contact these resources” (Knight et al., 2000:769). In two LONGSCAN studies, interviewers consult clinical staff if children appear distressed. In another study, interviews are conducted in schools and distressed children are asked to identify a counselor or teacher with whom they can discuss their problems.
Each LONGSCAN study has identified interview responses that could indicate abuse or neglect. Two LONGSCAN study teams report cases in which children give responses deemed to indicate they are at significant risk. Most interesting and controversial is the approach of two other LONGSCAN studies to a child’s indications of possible maltreatment. When a child responds in this way, the computer asks whether the child wishes to discuss possible abuse with the interviewer or another person. If the child answers no, the interview ends without further inquiry into the threat. The interviewer is unaware of the child’s responses, and the data are recorded in a manner that prevents other researchers from identifying the child (Kotch, 2000). Thus, the research team never obtains the information necessary to evaluate whether the situation should be reported. If the child
answers yes, the interviewer or an adult the child designates asks follow-up questions to determine whether reportable abuse has occurred. If it has, the child is told how and why a report will be made. If the problem does not appear to be reportable abuse, the child meets with an appropriate adult to discuss personal problems and receives information on local resources and how to obtain future assistance (Kotch, 2000).
In commenting on this approach, ethicist N. King described its benefits and risks to children and their families. The approach protects children and families from the harm that could come from a more liberal reporting approach, including unjustified disruptions in home life and an inadequate or damaging social services intervention. At the same time, the two LONGSCAN studies’ reporting model “privileges research over intervention,” on the assumption that better data will produce better interventions in the future. Yet researchers may be biased in making this determination, she noted. Furthermore, King questioned whether 12-year-olds should be put in the position of deciding whether to discuss their situations with researchers. She asked, “is this a means for adolescents to adopt or respect the values and priorities within their families, or is it a way for LONGSCAN researchers to pass the buck to their adolescent subjects?” (1999:184).
The above LONGSCAN approach could be adapted to elder abuse and neglect studies involving older adults with full or mildly compromised decisional abilities. Investigators could devise measures that enabled them to avoid learning about indications of possible abuse or neglect when study participants refused to discuss these matters. Investigators proposing such a strategy could face opposition from IRBs and others evaluating the ethics of the research, however.
Related issues are presented when elder abuse and neglect research involves surveys or other methods in which participants remain anonymous or research team members are blinded to participants’ responses. These approaches may be attractive because they allow investigators to guarantee confidentiality and avoid the need to devise measures to address suspected abuse and neglect. At the same time, such approaches have been challenged on ethical grounds. For example, the NBAC criticized the Common Rule’s failure to cover anonymous surveys that ask sensitive questions (2001:37). Similarly, ethicists N. King and Larry Churchill raised the following questions about the use of investigator blinding in interview studies of child abuse and neglect: “Might such a research design increase the risk of wrongs or harms to the child subjects who expect to develop some relationship with the researcher who asks such intimate questions? Will anonymity obviate researchers’ feelings of relationship with and obligations toward child subjects or just leave them with information on which they are powerless to act?” (2000:722).
The beneficence principle would support offering assistance to partici-
pants in research incorporating participant anonymity or investigator blinding. At minimum, investigators could supply to all prospective participants information on available health and social services for families and older persons. Research findings on the risk and prevalence of elder abuse and neglect in particular areas could also be shared with local protective services agencies and government officials, with the aim of increasing awareness of and services for at-risk individuals. Moreover, researchers should consider whether ethical principles would in some cases support alternatives to anonymity and blinding, alternatives such as “a thorough, thoughtful, nondeceptive, informed consent process in an ongoing research relationship, and a commitment to honor confidentiality unless grave harm appears imminent.” (King and Churchill, 2000:722).
Conflicts Regarding Information Disclosure
The second major conflict pits the interests of participants in being informed of research risks against societal interests in advancing knowledge. Sometimes withholding information from prospective participants would allow investigators to collect the most accurate data. Two such situations may arise. One occurs when investigators studying relatives or professional caregivers plan to report or take other actions if a research team member comes to suspect elder abuse or neglect. In this case, the concern is that prospective participants aware of the investigators’ plans will either refuse to enroll or if they do enroll, will provide inaccurate information. The other situation occurs when investigators are studying older persons. Here, the fear is that if investigators tell prospective participants (or their representatives) about plans to address suspected abuse or neglect, individuals opposed to protective interventions will refuse to enroll or, if they do enroll, will provide inaccurate information.
As indicated above, disclosure in both situations will often be required. The Common Rule permits exceptions to disclosure only when research presents no more than minimal risk to participants and when disclosure would make research impracticable. Underlying the narrowness of the exception is the ethical judgment that potential knowledge gains fail to justify depriving individuals of the freedom to refuse participation in research that exposes them to risks greater than those encountered in everyday life and routine health and social services practice (Beauchamp, 1996). Although failure to disclose reporting plans may be justified in clinical and social services programs designed to assist specific older persons, current policy deems such concealment unacceptable when data are collected for the purpose of advancing knowledge.
Current policy also recognizes that an investigator’s failure to disclose plans for addressing suspicions of neglect and abuse could have negative
long-term consequences for research, as well as the health care and social services systems. This is because research participants and community members who learn of the failure to disclose could lose trust in researchers, clinicians, and social workers (Bok, 1992).
At the same time, it should be recognized that the current policy requirements for disclosure could have negative consequences. The price of disclosure is a possible loss of societal benefits. This loss could occur if the disclosure requirements limit researchers’ abilities to assess and understand the problem of elder abuse and neglect. Whether this occurs will depend in part on whether researchers are able to devise methods that avoid or compensate for possible underreporting and other problems related to the disclosure requirements.
The research ethics literature fails to address information disclosure in the specific context of research on elder abuse and neglect. This topic has been covered, however, in articles examining research on other forms of harmful behavior. Discussions of the ethics of research on child abuse and neglect generally support disclosure of investigators’ plans to address suspected misconduct. For example, psychologist Celia Fisher argues for disclosure of all risks that could affect a child’s and parent’s willingness to enroll, including the possibility that suspected abuse will be reported (Fisher, 1999).
In the LONGSCAN studies of child abuse and neglect, investigators and reviewing IRBs favored disclosure as well. In these studies, parents, as well as children aged eight and older, are told that investigators will take various measures in response to suspected maltreatment, including reporting potentially serious harm to authorities. Parents also are given the option to review the questions their children will be asked. This represents a decision to give children and parents control over the information they provide, even though this may bias study results (Runyan, 1999). Study teams and IRBs in different LONGSCAN projects adopted different language to disclose reporting plans, however, with some disclosures more explicit than others. The language differences reflected differences in study responses to suspected abuse and neglect, which were discussed above.
A similar approach was adopted in the MacArthur Risk Assessment study of risk factors for community violence in former mental patients. In study consent forms, prospective participants were told that confidentiality would be preserved unless researchers believed the lives of participants or third parties were in danger or child abuse was discovered (Monahan et al., 1993). Thus, individuals had the option to refuse participation or to modify their response to investigators based on their awareness that reporting was possible.
In sum, if research participation will expose older persons or caregivers to risks greater than those present in ordinary encounters with health and
social services personnel, both research policy and ethics support disclosure of such risks. Balancing the goals of advancing knowledge, respecting informed decision making, and protecting research participants from harm will never be easy. The investigator’s overriding ethical obligation is to detect and evaluate potential conflicts in the initial phase of study planning. The appropriate ways to address conflicts will vary depending on the specific characteristics of the study. Often, there will be more than one ethically defensible option. What is most important is to develop approaches that are supported by a reasonable ethical analysis.
CONCLUSION: EXPANDING THE DELIBERATIONS
The Belmont Report principles and Common Rule requirements supply guidance on the appropriate conduct of research on elder abuse and neglect. Yet these principles and regulations are by necessity somewhat general. Every human research proposal presents its own issues; thus, careful deliberation by research teams, IRBs, and funding agency officials will always be essential.
These deliberations could be enriched by the views of people familiar with the actual situations of older persons, their families, and their caregivers. Since the Belmont Report and Common Rule were written, community consultation and representation have emerged as methods for making research more ethical. Indeed, the NBAC specifically advised that representatives of prospective research participants, including those from vulnerable populations, “should be encouraged to participate in the study design and oversight processes” (2001:91). By seeking information from members of the population of prospective research participants, investigators may learn of better ways to balance competing interests in their studies. As Celia Fisher and her colleagues have shown, people affected by research may have unexpected beliefs and attitudes about researchers’ responsibilities to participants.
Elder abuse and neglect research seems especially suited to this approach. Several writers have endorsed the involvement of a panel of residents and others who will be affected by research to review proposals to conduct research in nursing homes (Sachs et al., 1993). This concept could be extended to support other forms of community involvement in research on elder abuse and neglect. Members of the relevant communities could help investigators determine the facts that should be disclosed to prospective research participants and create effective methods for conveying the information. They could help researchers discern the appropriate balance between protecting participants’ privacy and preventing harm to vulnerable older persons. They could join researchers and social services workers in
developing appropriate responses to suspected abuse and neglect discovered in the course of a study.
Awareness of the basic ethical and policy considerations, combined with insights from the community of research participants, will supply the most solid moral foundation for research on elder abuse and neglect. A research program built on this foundation is most likely to receive continued public support, as well. The need for increased knowledge about this serious social problem is great, but so is the need to respect and protect the individuals, families, and communities affected by research.
Advisory Work Group on Human Subject Research Involving the Protected Classes 1998 Recommendations on the Oversight of Human Subject Research Involving the Protected Classes. New York: New York State Department of Health.
Amaya-Jackson, L., R.R.S. Socolar, W.M.Hunter, and D.K. Runyan 2000 Directly questioning children and adolescents about maltreatment. Journal of Interpersonal Violence 15:725–759.
Amdur, R., M. Speers, and E. Bankert 2000 IRB triage of projects that involve medical record review. IRB: A Review of Human Subjects Research 22(January-February 2000):4–7.
Angell, M. 2000 Is academic medicine for sale? New England Journal of Medicine 342:1516-1518.
Appelbaum, P., and T. Grisso 1998 Assessing patients’ capacities to consent to treatment. New England Journal of Medicine 319:1635–1638.
Appelbaum, P.S., and A. Rosenbaum 1989 Tarasoff and the researcher: Does the duty to protect apply in the research setting? American Psychologist 44(6):885–894
Beauchamp, T. 1996 Moral foundations. In Ethics and Epidemiology, S. Coughlin and T. Beauchamp, eds. New York: Oxford University Press.
Bellin, E., and N. Dubler 2001 The quality improvement-research divide and the need for external oversight. American Journal of Public Health 91:1512–1517.
Berg, J.W. 1996 Legal and ethical complexities of consent with cognitively impaired research subjects: Proposed guidelines. Journal of Law, Medicine and Ethics 24:18–35.
Bok, S. 1992 Informed consent in tests of patient reliability. Journal of the American Medical Association 267:1118–1119.
Bonnie, R. 1997 Research with cognitively impaired subjects. Archives of General Psychiatry 54:105–111.
Cassarett, D., J. Karlawish, and J. Sugarman 2000 Determining when quality improvement initiatives should be considered research. Journal of the American Medical Association 283:2275–2280.
Centers for Disease Control and Prevention 1999 Guidelines for Defining Public Health Research and Public Health Non-Research. Available: http://www.cdc.gov/od/ads/opspoll1.htm [Accessed November 28, 2001].
Dresser, R. 1996 Mentally disabled research subjects: The enduring policy issues. Journal of the American Medical Association 276:67–72.
1999 Research involving persons with mental disabilities: A review of policy issues and proposals. In Research Involving Persons with Mental Disorders That May Affect Decisionmaking Capacity, vol. 2. Bethesda, MD: National Bioethics Advisory Commission.
2001a Dementia research: Ethics and policy for the twenty-first century. Georgia Law Review 35:661–690.
2001b When Science Offers Salvation: Patient Advocacy and Research Ethics. New York: Oxford University Press.
Fisher, C. 1999 Relational ethics and research with vulnerable populations. In Research Involving Persons with Mental Disorders That May Affect Decisionmaking Capacity, vol. 2. Bethesda, MD: National Bioethics Advisory Commission.
Fisher, C., A. Higgins-D’Alessandro, J.B. Rau, T.L. Kuther, and S. Belanger 1996 Referring and reporting research participants at risk: Views from urban adolescents. Child Development 67:2086–2100.
Garfield, A. 1991 Elder abuse and the states’ adult protective services response: Time for a change in California. Hastings Law Journal 42:861–937.
Glantz, L, 1992 The influence of the Nuremberg Code on U.S. statutes and regulations. In The Nazi Doctors and the Nuremberg Code, G. Annas and M. Grodin, eds. New York: Oxford University Press.
Hastings Center 1987 Guidelines in the Termination of Life-Sustaining Treatment in the Care of the Dying. Bloomington: Indiana University Press.
High, D., P.J. Whitehouse, S.G. Post, and L. Berg 1994 Guidelines for addressing ethical and legal issues in Alzheimer disease research: A position paper. Alzheimer Disease and Associated Disorders 4:66–74.
Hoffman, D., and J. Schwartz 1998 Proxy consent to participation of the decisionally impaired in medical research— Maryland’s Policy Initiative. Journal of Health Care Law & Policy 1:123–153.
Institute of Medicine 1994 Women and Health Research: Ethical and Legal Issues of Including Women in Clinical Studies, vol. 1. A. Mastroianni, R. Faden, and D. Federman, eds. Washington, DC: National Academy Press.
Kass, N., and J. Sugarman 1996 Are research subjects adequately protected? A review and discussion of studies conducted by the Advisory Committee on Human Radiation Experiments. Kennedy Institute of Ethics Journal 6:271–282.
Keyserlingk, E.W., K. Glass, S. Kogant, and S. Gauthier 1995 Proposed guidelines for the participation of persons with dementia as research subjects. Perspectives in Biology and Medicine 38:319–362.
King, N. 1995 Privacy and confidentiality in research. In Encyclopedia of Bioethics, W. Reich, ed. New York: MacMillan.
1999 Research in distressed families. In Beyond Regulations, N. King, G. Henderson, and J. Stein, eds. Chapel Hill, NC: University of North Carolina Press.
2000 Defining and describing benefit appropriately in clinical trials. Journal of Law, Medicine & Ethics 28:332–343.
King, N., and L. Churchill 2000 Ethical principles guiding research on child and adolescent subjects. Journal of Interpersonal Violence 15:710–724.
Knight, E., D. Runyan, H. Dubowitz, C. Brandford, J. Kotch, A. Litrownik, and W. Hunter 2000 Methodological and ethical challenges associated with child self-report of maltreatment. Journal of Interpersonal Violence 15:760–775.
Kopelman, L. 1995 Research Policy: Risk and Vulnerable Groups. In Encyclopedia of Bioethics, W. Reich, ed. New York: MacMillan.
Kotch, J. 2000 Ethical issues in longitudinal child maltreatment research. Journal of Interpersonal Violence 15:696–709.
Monahan, J., P.S. Appelbaum, E. Mulvey, P.C. Robbins, and C. Lidz 1993 Ethical and legal duties in conducting research on violence: Lessons from the MacArthur Risk Assessment Study. Violence and Victims 8:387–396.
National Bioethics Advisory Commission (NBAC) 1998 Research Involving Persons with Mental Disorders That May Impair Decisionmaking Capacity, vol. 1. Bethesda, MD: National Bioethics Advisory Commission.
2001 Ethical and Policy Issues in Research Involving Human Participants: Report and Recommendations, Bethesda, MD: National Bioethics Advisory Commission.
National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research 1975 Research Involving Children: Report and Recommendations, DHEW Publication No. (OS) 76-127. Washington, DC: U.S. Government Printing Office.
1978 Research Involving Those Institutionalized as Mentally Infirm: Report and Recommendations. Washington, DC: U.S. Government Printing Office.
1979 The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Washington, DC: U.S. Government Printing Office.
National Human Research Protections Advisory Committee 2001 Children’s Workgroup Report Draft. Available: http://ohrp.osophs.dhhs.gov [Accessed December 5, 2001].
Office of Protection from Research Risks 1993 Protecting Human Research Subjects: Institutional Review Board Guidebook. Washington, DC: U.S. Government Printing Office.
O’Sullivan, C., and C. Fisher 1997 The effect of confidentiality and reporting procedures on parent-child agreement to participate in adolescent risk research. Applied Developmental Science 1:185– 197 .
Pritchard, I. 2000 Searching for research involving human subjects. IRB: A Review of Human Subjects Research 22(January-February):4–7.
Ratzan, R. 1985 Technical aspects of obtaining informed consent from persons with senile dementia of the Alzheimer type. In Alzheimer’s Dementia: Dilemmas in Clinical Research, V. Melnick and N. Dubler, eds. Clifton, NJ: Humana Press.
Runyan, D. 1999 Maltreatment in families: A research dilemma. In Beyond Regulations, N. King, G. Henderson, and J. Stein, eds. Chapel Hill: University of North Carolina Press.
2000 The ethical, legal, and methodological implications of directly asking children about abuse. Journal of Interpersonal Violence 15:675–681.
Sachs, G., and C. Cassel 1990 Biomedical research involving older human subjects. Law, Medicine & Health Care 18:234–243.
Sachs, G., J. Rhymes, and C. Cassel 1993 Biomedical and behavioral research in nursing homes: Guidelines for ethical investigations. Journal of the American Geriatrics Society 41:771–777.
U.S. Department of Health and Human Services 1981 Final Regulations Amending Basic HHS Policy for the Protection of Human Subjects. Federal Register 46:8366–8391.
1991 Federal Policy for the Protection of Human Subjects (Common Rule). Federal Register 56:28,012–28,018.
1999 Additional Protections for Children Involved as Subjects in Research, Code of Federal Regulations 45:§ 46.408(c).
U.S. Food and Drug Administration 1999 Protection of Human Subjects. Code of Federal Regulations 21: Parts 50 and 56.
Wendler, D., and K. Prasad 2001 Core safeguards for clinical research with adults who are unable to consent. Annals of Internal Medicine 135:514–523.
World Medical Association 2000 Declaration of Helsinki: Ethical principles for medical research involving human subjects. Journal of the American Medical Association 284:3043–3045.