Research
Ethical Issues in Research with Vulnerable Populations
Summary: This project attempts to determine which groups qualify as vulnerable, assess the adequacy of existing safeguards for clinical research with vulnerable populations, and identify additional mechanisms to ensure vulnerable populations receive appropriate protection.
| Section: | Ethics of Human Subjects Research | ||||||||||||||
| Principal Investigators: | David Wendler, Ph.D. Christine Grady, Ph.D. |
||||||||||||||
| Collaborators: |
|
Background: Vulnerability in research involves a diminished ability to protect one’s own interests, typically manifested through an inability to give informed or voluntary consent. Most children are unable to give their own voluntary, informed consent, as are some adults for various reasons. Federal regulations include specific additional requirements for research with children, prisoners, and pregnant women, and call for additional safeguards to protect vulnerable adults.
I. Research with Children: It is estimated that approximately 70 percent of medical treatments used in children have not been tested in children, even for basic safety and efficacy. This lack of evidence provides an important impetus to pursue pediatric research. A number of recent initiatives, including by the NIH and FDA, attempt to encourage and in some cases require pediatric research. With the expected increase in pediatric research comes increased emphasis on the need to ensure the children who participate in clinical research receive appropriate protection.
The Department’s efforts with regard to research on children began with a survey of 188 IRB chairpersons around the country (Shah 2004). The federal regulations direct IRBs to categorize the risks of pediatric research in comparison to the risks children face in daily life. Specifically, a research intervention qualifies as minimal risk provided its risks do not exceed the risks children face in daily life or during the performance of routine procedures. Our survey found wide variation in how IRB chairpersons apply this minimal risk standard with, for example, 48 percent of chairpersons categorizing an MRI as minimal risk and 44 percent categorizing the very same procedure as more than minimal risk. We speculated that, in the absence of systematic data on the risks children face in daily life, IRB chairpersons might be relying on personal judgment to assess the risks of pediatric research. Extensive psychological research establishes, however, that reliance on personal judgment or intuition, in the absence of systematic data, is an unreliable method of risk assessment.
In an effort to develop systematic data, we followed this survey with an extensive review of existing data bases on the risks children face in daily life (Wendler 2005). These data provide IRBs with a baseline for systematically assessing whether research risks exceed the risks children face in daily life. Because the risks children face in daily life often are of different types compared to the risks posed by research procedures, IRBs cannot simply compare the risks of research participation directly to the risks of daily life. For example, airway intubation as part of a research study poses some risk of airway abrasion whereas daily life poses essentially no risk of airway abrasion to ordinary children. To facilitate implementation of the minimal risk standard, we developed a systematic method, which we call ‘comparative’ analysis, for comparing different types of risks (Wendler, Varma 2006). This approach divides the potential harms children face in daily life into five levels of severity, from negligible to catastrophic, and quantifies the chances that children will experience a harm at each respective levels from the activities of daily life. This approach provides IRBs with a systematic method to determine whether research procedures pose greater than minimal risk.
Our initial survey evaluated how IRB chairpersons apply the federal minimal risk standard to hypothetical examples. Development of the comparative analysis method put us in a position to address this limitation by assessing how IRBs apply the minimal risk standard in practice, to actual clinical research studies. In particular, we applied the comparative analysis method to studies submitted to OHRP for possible approval in category 407 of the regulations (Wendler, Varma 2006). This analysis suggests misapplication of the minimal risk standard is widespread, underscoring the need to develop systematic methods, such as the comparative analysis, and assess them in practice, for implementing the federal minimal risk standard.
The federal regulations allow children to be enrolled in research that poses greater than minimal risk and offers no prospect for direct benefit when the risks do not exceed a ‘minor’ increase over minimal risk. Although this standard has been in place for over 20 years, there have been few systematic attempts to define what constitutes a minor increase over minimal risk. We have argued that the most widely cited approach, endorsed by Weijer and Freedman, suffers from a number of fatal flaws (Wendler, Emanuel 2005). Most importantly, this approach is limited to psychological risks. We proposed instead that a ‘minor increase’ over minimal risk should be understood as referring to the risks in the daily lives of children who face greater than average, but nonetheless socially acceptable risks, such as the risks faced by children who work on family farms.
Our survey of IRB chairpersons also revealed wide variation in implementation of the “assent” requirement, the only requirement in the federal regulations to address the extent to which children should have a say in whether they are enrolled in research. Previously, it had been widely assumed that most IRBs regard children as being capable of assent at the age of 7, a threshold that has been endorsed by the National Commission and American Academy of Pediatrics. Our survey found IRBs use a range of ages from as low as 5 years to as high as 14 years (Whittle 2004). We also found that almost half of IRBs do not use an age cutoff at all, but leave this determination to investigators' clinical judgment.
These data highlighted the need to develop systematic guidelines on the age at which children are capable of assent. For this purpose, we considered first the purpose of soliciting the assent of children. This analysis suggests children should be deemed capable of assent when they can make their own decisions regarding research participation. We evaluated existing data on cognitive development. On this basis, we argued that the assent threshold should be set at approximately age 14 (Wendler, Shah 2003). This analysis also suggested that it is a mistake to rely exclusively on the assent requirement for pediatric research since this requirement does not provide any reason to respect the dissent of children deemed incapable of assent. To address this concern, we have argued that investigators also should respect the sustained dissent of all children.
Impact of Research: As a result of its work on pediatric research ethics, consultation from members of the department has been requested by two IOM committees working on pediatric research ethics. The department’s work on minimal risk was cited in the IOM committee report on pediatric clinical research. The department also was asked to write the principal background paper for the report of the IOM committee on housing research with children. The department's work on minimal risk and assent have been widely cited and presented at numerous conferences.
Current and Future Initiatives: The federal regulations on pediatric research, like many guidelines around the world, focus on ‘direct’ benefits. On this approach, the risks to which children are exposed in the research context can be compensated by direct benefits, but not by indirect benefits. Unfortunately, there has been little examination of this crucial concept. The most systematic analysis comes from Nancy King, who defines a direct benefit as one that is the result of the intervention being tested in the study in question. We are currently working on an analysis which suggests King's account is unnecessarily restrictive. Direct benefits are better understood as benefits that result from the interventions necessary to answer the scientific question posed by the protocol.
We also are developing a practical method IRBs can use to assess the risks of pediatric research. The need for such a method was highlighted by our survey which found that many IRB chairpersons do not understand what constitutes a “prospect” of direct benefit, while others were implementing this requirement inappropriately, assuming that any chance of direct benefit was sufficient for a protocol to qualify as offering a prospect of direct benefit. Finally, recent media coverage has focused on the ethical issues concerned with enrolling in research children who are wards of the state. The department has undertaken an analysis of the present regulations for wards, which suggests several modifications should be adopted to ensure appropriate protection for this especially vulnerable population. In particular, the scope of the responsibility of advocates should be clarified and expanded, and wards should be enrolled in research only when there is a compelling scientific reason why the study needs to enroll them.
II. Research with Vulnerable Adults: Many groups of adults have been described as vulnerable, to the extent that it can be difficult to determine who is not vulnerable, thereby diminishing the possibility of meaningful additional protections. To take one example, guidelines published by the Council of International Organizations of Medical Sciences (CIOMS) include junior or subordinate members of a hierarchical group, such as “medical and nursing students, subordinate hospital and laboratory personnel, employees of pharmaceutical companies, and members of the armed forces or police...the elderly...residents of nursing homes, people receiving welfare benefits or social assistance and other poor people and the unemployed, people in emergency rooms, some ethnic and racial minority groups, homeless persons, nomads, refugees or displaced persons, prisoners, patients with incurable disease, individuals who are politically powerless, and members of communities unfamiliar with modern medical concepts” as vulnerable.
In addition to including too many groups, current concepts of vulnerability are applied to whole groups of people, without distinguishing between individuals in a group who might truly have compromised capacity to protect their own interests from those who do not. Considering all poor people, pregnant women, members of ethnic or racial minorities, and people with terminal illness as inherently vulnerable has been particularly controversial. An individual's needs for special protections in the research context may depend more on personal factors, characteristics of the research and the research environment, available alternatives, than merely on that person’s membership in a particular sociodemographic group.
The federal regulations for clinical research stipulate that IRBs should include additional protections for individuals who are deemed vulnerable, or exclude them. Yet, there has been little analysis or recommendation in the literature on what constitutes appropriate protections in this regard. The existing regulations provide additional explicit requirements for three groups: prisoners, pregnant women and fetuses, and children. However, the National Commission's recommendations for research with cognitively impaired adults were never adopted. This gap is especially troubling because existing regulations rely heavily on the protections afforded by informed consent, possibly leaving adults who cannot consent without adequate safeguards.
The Department's work in this area began in collaboration with members of the Consortium to Examine Clinical Research Ethics (CECRE established in 2001 by the Doris Duke Charitable Foundation) examining the concept of vulnerability and critiquing current understandings and use of the concept (Levine et al. AJOB 2004). We argued that with a few exceptions (such as children) instead of labeling whole groups of people vulnerable, other aspects of research such as the particular design, the inherent risks, the level of uncertainty, etc. should warrant ‘special scrutiny’ (Levine et al. Annals 2004).
Building on this work, we have undertaken several projects related to understanding vulnerability of the socioeconomically disadvantaged in research. Members of the department undertook an analysis of the inclusion in research of the uninsured, arguing that categorical exclusion was unfair but that targeting may also be unfair. Working with NIH collaborators from NIAMS and the NIH Clinical Center, we began to explore issues related to vulnerability in the uninsured and economically disadvantaged with the core community group of the Health Partnership Program formed by NIAMS and the Upper Cardozo Clinic in Washington DC. Using vignettes, we explored issues related to exploitation, vulnerability and trust with this core group (Grady et al. 2006). After a small series of individual interviews and additional meetings with the HPP core community group, we are beginning an empirical study of research participants at the Cardozo clinic to better understand their views about research and vulnerability.
Concomitantly, we examined common arguments and concerns about vulnerability of the economically disadvantaged, concluding that there is little justification for routinely excluding the economically disadvantaged from research or categorically considering them vulnerable (Denny and Grady, In press).
It has been widely assumed that individuals from minority groups in the United States, especially African Americans, are less willing to participate in health research compared to non-Hispanic whites. The literature describes this lack of willingness as based on prevalent distrust among African Americans in research and the research establishment, evolving from a legacy of abuse. This assumption has led many to assume that any efforts to increase minority participation in research must first focus on individuals’ attitudes. Despite the prevalence of this view, there has been no systematic assessment of whether individuals from minority groups are less willing to participate, when eligible and invited. We conducted a systematic analysis of the published literature to assess this view. Our analysis suggests, contrary to the common view, that individuals from minority groups are as willing as whites to participate in health research (Wendler et al. 2006).
The Department's work on identifying appropriate protections for vulnerable individuals began by arguing that enrollment in research of individuals who are in fact unable to provide informed consent poses increased moral risks, compared to the enrollment of individuals who can consent (Wendler 1998, Wendler 2000). By analogy to individuals who face increased physical risks, for example, individuals with decreased kidney function, this conclusion implies that individuals who cannot consent should not be enrolled unless there are compelling reasons to enroll them. We have labeled this safeguard the “necessity” requirement.
There has been much debate over whether adults who cannot consent should be enrolled in research that offers them the prospect of direct benefit. Some argue that such a restriction represents discrimination against a vulnerable group; others argue that these individuals deserve additional protection in the research setting. Our analysis of the necessity requirement suggested that these limitations are justified not on the grounds that they protect the interests of specific individuals, but that they minimize the aggregate risks of the research. This analysis suggests individuals who cannot consent should not be enrolled unless necessary, even when the research offers them the prospect of direct benefit.
We next conducted a systematic assessment of the existing recommendations for research with cognitively impaired adults. Based on this work, we developed a set of core safeguards for this type of research (Wendler, Prasad 2001). In this work, we argued that the necessity requirement provides better and more appropriate protection than the widely endorsed “subject's condition” requirement.
One protection that has been advocated by a number of groups is the requirement that adults who cannot consent should be enrolled in research only when they have completed a formal research advance directive. To assess this requirement, we evaluated the attitudes of participants in research studying individuals at increased risk of developing Alzheimer disease (Wendler 2002). The results suggested that, in the absence of a formal advance directive, these individuals were willing to allow their family members to decide whether to enroll them in research should they lose the ability to make these decisions for themselves.
This survey also attempted to assess the completion rate of research advance directives by measuring how many individuals completed a research advance directive when it was provided to them. The respondents had just completed a survey of the importance of research advance directives, and the vast majority expressed support for their use. Yet, less than one in five actually completed a research advance directive. These data may not track completion rates in clinical practice. To address this limitation, we assessed research participants who come to the NIH Clinical Center which enquires whether all inpatients have an advance directive, provides written information on advance directives and encourages individuals to complete an advance directive. We found that only 11 percent of inpatients completed a research advance directive (Muthappanian 2005). These data provide further evidence that requiring a formal advance directive for adults who cannot consent for themselves would block many individuals from participating in research, including those who had expressed an interest in such research and were willing to allow their family member to enroll them. We conclude that a research advance directive should be understood as a recommendation, not a requirement and in the absence of a formal advance directive, family members should be allowed to make research decisions. In addition, to implement the substituted judgment standard, as the risk/benefit profile of research becomes less favorable to participants, more evidence should be required of their own preferences about participation. .
Impact of Research: The Department’s work on research with vulnerable adults, adults who cannot consent for themselves, and the role of advance directives in such research has been widely cited. We have received numerous requests from institutions and IRBs around the country, who use our proposed guidelines on research with adults who cannot consent to inform their policies and practices. Since it publication this year, our assessment of minorities’ willingness to participate in research has generated a great deal of interest, including presentation at a DHHS conference on research with minority populations. This work was also the centerpiece for an office of minority health workshop on increasing minority participation in clinical research.
Current and Future Initiatives: The NIH Clinical Center was one of the first research institutions in the country to adopt an explicit policy regulating research with cognitively impaired adults (MAS 87-4). Since adoption of the policy almost 20 years ago there has been an evolution in thinking on the topic. The department is currently working with the Clinical Center's ethics committee to develop a revision of this policy to reflect in practice the recommendations we have made in this area. Second, we have been conducting research on decision making by surrogates. This work reveals that almost all the work to date in this area has been in the context of clinical care. Future studies will assess the nature of surrogate decision making in the research setting. Research participants at the Cardozo clinic will be interviewed in Spanish or English to ascertain their views about research, vulnerability, exploitation, and related concepts. The department plans to continue its work examining the meaning and limitations of the concept of vulnerability, especially in economically disadvantaged populations both domestically and internationally.
Publications:
Wendler D, Varma S. Improving Pediatric Medicine Through Research: Current Obstacles, Proposed Solution. Journal of Pediatrics. In press.
Grady C, Hampson L, Wallen G, Rivera-Goba M, Carrington K, Mittleman B. Exploring Perceptions about the Ethics of Clinical Research in an Urban Community. American Journal of Public Health. 2006:96(11):1996-2001.
Denny C, Grady C. Clinical Research and the Economically Disadvantaged: Clarifying Concerns. J Med Ethics. In press.
Wendler D, Cornelio M. Overcoming language barriers in medical care. Pediatric Blood and Cancer. In press.
Wendler D, Miller F. Assessing research risks systematically: the net risks test. Journal of Medical Ethics. In press.
Peerzada JM, Wendler D. Hematopoietic stem cell transplant research with pediatric donors: when can institutional review boards approve it? Transplantation. 2006;81:1616-1620.
Wendler D. Assent in paediatric research: theoretical and practical considerations. Journal of Medical Ethics. 2006;32:229-234.
Wendler D, Kington R, Madans J, Van Wye Heidi G, Christ-Schmidt H, Pratt LA, Brawley OW, Gross CP, Emanuel EJ. Are racial and ethnic minorities less willing to participate in health research? PLoS Medicine. 2006;3:e19.
Buchanan DR, Miller FG. Justice and Fairness in the Keendy Krieger Institute Lead Paint Study: the Ethics of Pulic Health Reearch on Less Expensive, Less Effective Inteventions. American Journal of Public Health. 2006;96:781-787.
Wendler D, Emanuel EJ. What is a ‘minor’ increase over minimal risk? Journal of Pediatrics. 2005;147: 575-578.
Wendler D, Belsky L, Thompson KM, Emanuel EJ. Quantifying the federal minimal risk standard: implications for pediatric research without a prospect of direct benefit. JAMA. 2005;294:826-832.
Lands LC, Allen J, Cloutier M, Leigh M, McColley S, Murphy T, Wilfond B. Pediatric Assembly of American Thoracic Society Subcommittee. ATS Consensus Statement: research opportunities and challenges in pediatric pulmonology. American Journal of Respiratory Critical Care Medicine. 2005;172:776-780.
Muthappanian G, Forster H, Wendler D. Research advance directives: protection or obstacle? American Journal of Psychiatry. 2005;162:2389-2391.
Wendler D. Protecting subjects who cannot give consent:toward a better standard for 'minimal' risks. Hastings Center Report. 2005;35:37-43.
Levine C, Faden R Grady C Hammerschmidt D, Eckenwiler L, Sugarman J of the Consortium to Examine Clinical Research Ethics. The Limitations of “Vulnerability” as a Protection for Human Research Participants. American Journal of Bioethics. 2004;4(3):44-49.
Levine C, Faden R, Grady C, Hammerschmidt D, Eckenwiler L, Sugarman J. on behalf of the Consortium to Examine Clinical Research Ethics. “Special Scrutiny”: A Targeted Form of Research Protocol Review. Annals of Internal Medicine. 2004;140:220-223.
Hurst, S. When Patients Refuse Assessment of Decision-Making Capacity: How Should Clinicians Respond? Archives of Internal Medicine 2004; 164:1757-1760.
Wendler D. Can we ensure all research subjects give valid consent? Archives of Internal Medicine. 2004;164:2201-2204.
Ulrich CM, Grady C, Wendler D. Palliative care: a supportive adjunct to pediatric phase I clinical trials for anti-cancer agents? Pediatrics. 2004;114:852-855.
Whittle A, Shah S, Wilfond B, Gensler G, Wendler D. IRB practices regarding assent in pediatric research. Pediatrics. 2004;113:1747-1752.
Wendler D. Risk standards for pediatric research: rethinking the Grimes ruling. Kennedy Institute of Ethics Journal. 2004;14:189-200.
Wendler D, Forster H. Why we need legal standards for pediatric research. Journal of Pediatrics. 2004;144:150-153.Shah S, Whittle A, Wilfond B, Gensler G, Wendler D. How do IRBs apply the federal risk and benefit standards for pediatric research? JAMA. 2004;291:476-482.
Taylor HA, Wilfond BS. Ethical issues in newborn screening research: lessons from the Wisconsin cystic fibrosis trial. Journal of Pediatrics. 2004;145:292-296
Coffey MJ, Wilfond B, Ross LF. Ethical assessment of clinical asthma trials including children subjects. Pediatrics. 2004;113:87-94.
Wendler D, Shah S. Should children decide whether they are enrolled in non-beneficial research? American Journal of Bioethics. 2003;3:1-7.
Wendler D, Shah S, Whittle A, Wilfond B. Non-beneficial research with individuals who cannot consent: is it ethically better to enroll healthy or affected individuals? IRB: Ethics and Human Research. 2003;25:1-4.Pace C, Miller FG, Danis M. Enrolling the Uninsured in Clinical Trials: An Ethcal Perspective. Critical Care Medicine. 2003;31:S121-S125.
Rosenstein DL, Miller FG. Ethical considerations in psychopharmacological research involving decisionally impaired subjects. Psychopharmacology. 2003;171:92-97.
Wendler D, Martinez R, Fairclough D, Sundlerand T, Emanuel EJ. Proposed regulations for clinical research with adults unable to consent: what are the views of those most likely affected? American Journal of Psychiatry. 2002;159:585-591.
Chen DT, Miller FG, Rosenstein, DL. Enrolling decisionally impaired adults in clinical research. Medical Care. 2002;40:v20-v29.
Wendler D, Prasad K. Core safeguards for clinical research with adults who are unable to consent. Annals of Internal Medicine. 2001;135:514-523.
Wendler D. Informed consent, exploitation, and whether it is possible to conduct human subjects research without either one. Bioethics. 2000;14:310-339.
Wendler D. The importance of autonomy not being all-important. BioLaw. 1999;S:445-451.
Jaworska, A. Respecting the Margins of Agency: Alzheimer’s Patients and the Capacity to Value. Philosophy and Public Affairs. 1999;28:105-138.
Miller F, Rosenstein D. Independent Capacity Assessment: A Critique. BioLaw Special Section. September-October 1999;S:432-439.
Wendler D. When should ‘riskier’ subjects be excluded from research? Kennedy Institute of Ethics Journal. 1998;8:307-327.
