Putting the "ethics" back into research ethics: a process for ethical reflection for human research protection.

Introduction

"When persons prevent their emotions from overtaking their rationality, it is called reason. When persons prevent their rationality from overtaking their emotions, it is called compassion. When persons can do both, it is called wisdom."

--Ancient Chinese Saying

Research involving human subjects should be a partnership between the subjects who volunteer for the research, the public who hope to benefit from future drugs, devices, and biological products, investigators who design and conduct research, sponsors (institutions and companies that design, conduct, and sponsor or fund research), and IRBs who review research. I argue that "compliance with the regulations" governing research with human subjects has taken a front seat to broader ethical considerations of such research. While regulations contain practical applications of the ethical principles respect for persons, beneficence, and justice from such documents as the Belmont Report and Declaration of Helsinki, regulations do not explicitly reference these principles or provide procedures for ensuring that those who design, conduct, or review research have properly considered such matters.

Compliance is important, but it can become too focused on "checking off boxes" and the bureaucracy of documentation. As Dr. Gregory Koski (2003), former Director of the Office for Human Research Protections (OHRP), articulated, human research protection programs (HRPPs) need to move from a "culture of compliance" to a "culture of conscience" (p. $5). I argue that part of this shift in institutional culture involves re-discovering the ethical underpinnings of regulations for the protection of human subjects in such documents as the Belmont Report, the Nuremberg Code, and the Declaration of Helsinki. It is not enough to merely mention these documents in multi-center clinical trial protocols or IRB policy and procedure documents. Rather, the ethical principles in these documents need to be brought to life or enlivened during discussions of how to properly design, conduct, or review research.

Unfortunately, the number of findings from regulatory inspections from the Food and Drug Administration (FDA) and OHRP seems to indicate that there is a fundamental lack of training on the ethical principles themselves. The findings from such inspections consistently point to lapses in informed consent procedures and lack of investigator knowledge of human subject protection requirements. In addition, regulations provide little procedural guidance as to what the ethical substance of an IRB meeting should be like. However, regulations do provide checklists against which to document the presence or absence of certain required elements in protocols, informed consent documents, and recruitment materials.

Without training or experience in how to bring ethical principles to life, those who design, conduct, or review human subjects research will likely focus on ensuring compliance with regulatory requirements. In this way, they can at least be sure that the bare minimum of requirements has been met. This paper represents an attempt to supply those involved in designing, conducting, or reviewing human subjects research with a process of enlivening ethical principles by relying on a pragmatic framework for fostering moral discourse both within our own consciences and with others involved in these three research activities.

Origins of the Ethical Principles of Human Subjects Research

In 1979, the National Commission for the Protection of Human Subjects in Biomedical and Behavioral Research (the National Commission) produced, arguably, its most influential report. The Belmont Report, so named because the members of the Commission met at the Smithsonian Institute's Belmont Conference Center, articulated the ethical principles and philosophical basis for conducting research involving human subjects. The Report espouses three distinct ethical principles--respect for persons, beneficence, and justice--then provides practical measures for how these principles should be implemented in the design, conduct, and review of research (National Commission, 1979). The Report discusses the philosophical foundations of each of these principles and why each principle should be used in the research context. Discussion of these finer philosophical points is beyond the scope of this particular paper, but a multitude of philosophical sources, ranging from Aristotle, Immanuel Kant, John Stuart Mill, and David Hume, among others, find their thinking contained in the Belmont Report. The writers of the Belmont Report specifically intended and hoped that the Report and the three ethical principles would serve both as guideposts and reminders for those involved in designing, conducting, or reviewing human subjects research.

An "Ethics Gap" Between the Belmont Report and its Regulatory Offspring

It is not known, beyond mere anecdote, whether the Belmont Report enjoys wide dissemination to investigators (either in academic or medical settings or those in industry) nor to what extent the three ethical principles are part of education and training given to such groups. Moreover, it is not known whether those involved in reviewing research have read the Belmont Report, despite its free availability via the Internet on the FDA and OHRP websites.

What the Belmont Report provides in terms of philosophical justification and ethical analysis, it lacks in specific procedures for how its readers might employ the principles in designing, conducting, or reviewing research from an ethical perspective. Of course, that was and is not the purpose of the Belmont Report or the intent of its writers. Much of the practical import of the Belmont Report focuses on the translation of the ethical principles into their respective applications (respect for persons into the requirement to obtain informed consent, beneficence into the requirement to balance risks and benefits, and justice into the requirement of equitable subject selection). Current FDA, Department of Health and Human Services (DHHS), and international regulations governing human subjects research include each of the practical translations of the three ethical principles from the Belmont Report, yet no explicit mention of the ethical principles themselves appear in any of these regulations. Thus, there is a gap between the philosophical and the regulatory. Regulations mandate minimum requirements, but lack the specificity of procedural guidelines for how those who design, conduct, or review research should apply the ethical principles in their work. That is, and should not be, the purpose of regulations, for how would a government agency, institution, or company measure compliance with such an abstract notion as "ethical awareness."

To summarize, there is a paucity of evidence that those involved in human subjects research are aware of and utilize the ethical principles from the Belmont Report in their daily work. Moreover, the regulations do not directly address or mention the ethical principles themselves. Given these two pieces of information, it is not clear that reliance on regulations alone to carry out the ethical work required for appropriately designing, reviewing, or conducting research with human involvement is sufficient to ensure that the intent behind the Belmont Report can be met.

This "ethics gap" between the Belmont Report and the regulations might result in any of the following deficiencies in those who design, conduct, or review human subjects research:

1. Lack of awareness of particular ethical problems in the research;

2. Lack of ability or knowledge of how to apply ethical principles (and thus lack of acceptable resolutions to ethical problems in research); and

3. Focus on regulatory compliance rather than the broader ethical context and issues raised by the design or conduct of the research itself.

Bridging the "Ethics Gap:"Accreditation and Beyond

A series of research tragedies and scandals in the past decade have resulted in increased governmental scrutiny and focus on human subject protection regulatory compliance in the United States, both from OHRP and FDA. As a result, companies and institutions have become more focused on regulatory compliance as well. This focus might even be responsible for widening the "ethics gap" between regulations and ethics.

Fortunately, recent efforts to accredit human research protection programs (HRPPs) and thus the move away from exclusive focus on regulatory compliance to broader considerations in human research protections should help narrow this widening gap. At the time of the writing of this manuscript, two groups were offering accreditation of HRPPs--the Association for the Accreditation of HRPPs (AAHRPP) and the Partnership for Human Research Protection (PHRP). Both groups produced standards by which they will assess HRPPs, and both sets of standards explicitly address the ethical training not only of investigators but also IRB members and staff.

However, much like the Belmont Report, which articulates the ethical ideals to which those involved in research should aspire, the accreditation standards do not supply specifics on the mechanisms that institutions and companies should use to meet the standards. Thus, the purpose behind this paper.

I have suggested that without a process or framework for how those involved in designing, conducting, or reviewing research can enliven the ethical principles in their daily work, compliance concerns will rule the day and the "ethics gap" will widen. The process or framework that will help bridge this "ethics gap" must by necessity reference both regulations and ethical principles, but what else ought it to include?

A very interesting literature is emerging in bioethics that suggests the process of ethical inquiry and evaluation begins with our initial reaction to whatever is proposed for our ethical consideration. Authors have given different names or labels to this process or elements in it--the "Yuck" factor, intuition (Fins, Bachetta, & Miller, 1997), conscience (Callahan, 1988 and 1991), or listening to one's "gut" feeling (Hayes, 1986; Wocial, 1996; Fletcher, Miller, & Spencer, 1998; Purtillo, 1999). Most authors agree, however, that this initial reaction requires careful consideration using facts, principles, rules, regulations, and other contextual aspects to the problem in order to reach the pinnacle of ethical inquiry and evaluation. While the work of these authors is primarily concerned with the resolution of moral dilemmas in clinical care settings, the frameworks used to supply clinicians and ethics consultants a process for assisting in the resolution of clinical issues might also serve to bridge the "ethics gap" in the protection of human subjects.

One of these approaches relies on the pragmatic philosophy of John Dewey. Termed clinical pragmatism, the goal of ethical inquiry is to tutor one's intuitions or feelings about a particular situation, and using communication skills, facts, and ethical principles or rules as hypotheses to be explored and validated during the conversation, reach consensus regarding the alternatives identified (Fins, Bachetta, & Miller, 1997).

The goal of clinical pragmatism certainly matches the hoped-for outcome when an IRB considers a protocol or when a company seeks ethics advice for a proposed protocol or research design. Nevertheless, consensus is not the hoped-for outcome when an individual investigator believes an ethical problem exists in the design or conduct of the actual research. Perhaps an analogous process, drawn from others in the pragmatist tradition, would better suit all three human subject protection contexts discussed in this paper.

Like Dewey, fellow pragmatist Henry David Aiken was concerned with the process of moral inquiry and discourse. Unlike Dewey, the end point of Aiken's levels of moral discourse, which I would argue are analogous to Dewey's process of inquiry, is not consensus. It does pertain to reaching a decision, but it does not posit a specific outcome. Rather, the goal of Aiken's fourth level of discourse involves asking and answering, for oneself and with others who share a particular moral problem, one question: "Why should I (we) be moral?" I maintain that asking and attempting to supply an answer to that question, in the context of working to protect human subjects and uphold their dignity, ultimately brings to life the question of ethics and reinforces the whole reason that research exists--to gain scientific knowledge while at the same time protecting those people who voluntarily assume the risks for such knowledge. The remainder of this paper will supply a summary of Aiken's work and how this work can be adapted as one remedy to the gap between ethics and regulations and inform the ethical work required of those who design, conduct, or review research involving human subjects.

Aiken's Four Levels of Moral Discourse and The Integrative Nature of Morality

Much of the history of philosophy has tended to either minimize the value of emotion in moral decision-making or suggest it negatively impacts the ability to make rational moral decisions. This received view has been intensely criticized lately, largely in part because of a renewed interest in the role of emotion in morality, and increased scientific understanding of the neurobiological bases of emotion. The combined evidence from these diverse areas of thought refutes the so-called "negativity" traditionally associated with emotions and their role in morality and posits emotion as an essential component of competent moral decision-making. In fact, much of the current thinking is that emotion is a necessary, key component of moral life--often its starting place. Several sources in philosophy and bioethics have suggested that the ability to feel conflicting emotions and obligations help us recognize the presence of a moral dilemma (Hayes, 1986; Sherman, 1990; Wocial, 1996; Fletcher, Miller, & Spencer, 1998; Purtillo, 1999). These conclusions are precisely the ones that philosopher Henry David Aiken drew decades before, when the "negativity" theory of emotion held philosophical sway among so-called cognitivists who attempted to reduce all moral judgments to our ability to rationally evaluate competing alternatives. Aiken believed that morality consisted in more than mere rationality.

To remedy the tendency of moral theory to be monistic and reduce all moral judgments to one governing principle or process of reasoning and exclude feeling and emotion from the moral realm, Aiken posits four distinct, procedural levels of moral discourse. The result of ignoring the integrative nature of morality, Aiken argued, is to misunderstand the role of several features of the moral life and put ethical discourse into interminable, irreconcilable, and profitless debate. Aiken's framework incorporates the features of moral life that he finds lacking in other theories.

First, Aiken argues that a moral framework must acknowledge both the cognitive and emotive elements in moral decision-making and avoid the tendency to focus merely on one aspect or the other as determinative of morality. The same holds for subjectivity and objectivity; moral discourse has elements of both, and our theories should recognize this fact. Moreover, moral discourse must acknowledge the importance of reason, but also recognize its limits.

Aiken equates the first level of moral discourse with our unreflective, gut reaction to situations. He terms this the expressive-evocative level. Expressions of pleasure and displeasure vent our emotional reactions to events or situations. Shock at receiving a serious medical diagnosis, fear and uncertainty when confronted with one's own mortality in battling a terminal illness, and a mother's joy after delivering a healthy child are examples of expressive-evocative responses.

Aiken argues that normally there is no need to evaluate or reflect on such expressions of emotion; they are just personal, emotional manifestations of our pleasure or displeasure, "the venting of contrary--or perhaps merely different--emotions" (Aiken, 1962, p. 69).

Thus, to seek justification or to ethically evaluate the inherent wrongness or rightness of an emotive reaction is either senseless or it immediately shifts moral discourse to an altogether different level. This shift in moral discourse shows Aiken understood that to morally evaluate an emotive reaction requires more cognitively sophisticated tools of moral discourse.

The shift from the expressive-evocative level to more cognitive levels of processing suggests that, for Aiken, emotion and cognition are connected in moral discourse. The second level of discourse, the moral level, involves seeking justification for the "goodness" or "badness" of our expressive-evocative statements in light of two factors. The first factor combines an appraisal of the facts of the matter, the means of achieving the various moral outcomes, and the consequences that result from each. The second factor, the search for the rules or procedures that establish the relevancy of our appraisals, provides initial justification for particular moral actions. Thus, our emotive reactions at the first level require factual appraisals and consideration of consequences and rules of thumb or guidelines that might fit the facts.

Given this view, it is nonsensical to base judgments about one's conduct or actions solely on the basis of our emotive reactions. Our emotions are often strong, producing equally strong outward expressions, and, if we allow them, such expressions can control our decisions. The result can often be harmful, especially if we are trying to make decisions with others such as those made by an IRB or other bodies charged with protecting research subjects. Fortunately, the fourth level in Aiken's model can counteract this tendency (whereas the third level, which will be discussed afterwards, concerns more cognitive processes).

Fletcher and colleagues describe moral ambiguity as the uniquely human ability to feel conflicting emotions and obligations (Fletcher, Miller, & Spencer, 1998). While moral ambiguity is normal, in morality "decision is king": moral questions require answers and action (Aiken, 1962, p. 87). Aiken frames the solution to moral ambiguity by asking the paradoxical question "Why should I be moral?" He terms this fourth level "the human level," because the ability to answer this question depends on the human capacity to be moved. Since the question concerns finding motivation for the human heart, Aiken argues:

   I am "satisfied" and the question is
   "answered" not when some objective
   conditions have been met but when my
   practical indecision or doubt has been
   removed.... Here the only sort of justification
   possible is of the subjective sort
   which provides an "exciting occasion"
   capable of motivating the will (Aiken,
   1962, p. 86).

Asking "Why be moral?" prompts examination of the motivations that human beings feel: to act morally, to express moral feelings, to define the terms of their moral rules, to seek justification for such rules, and finally to avoid lapsing into ambivalence or indifference regarding the essential practical character of moral life. In this sense, the question "Why should I be moral?" provides the human heart reasons for continued moral discourse and action when reason and emotion fail to motivate the will (Fletcher, 1998, personal communication). Thus, the ability to reach an ethically justifiable resolution requires more than the ability to bring rational tools of ethical analysis or rules and regulations alone to bear on the essentially practical nature of moral questions. If the decision to be reached involves others, as decisions in IRB meetings or other bodies for protecting research subjects often do, then maintaining respect and concern for others--seeing them as persons worthy of respect either as fellow participants in the moral dialogue or the object of such dialogue (the human subjects on whose behalf moral judgments are made regarding research)--becomes all the more important for remaining engaged in moral dialogue.

The first and last levels of Aiken's framework provide the subjective, emotive elements of moral discourse. To counterbalance these, Aiken's second and third levels include cognitive, reasoning elements. The second, moral level of discourse concerns the rational evaluation and search for the rules (and regulations) that justify our emotive reactions at the first level. Although normally most moral discourse need not proceed beyond the moral level, Aiken recognized that sometimes the rules can be in direct conflict or the options for action reached might be unethical. When such situations confront us, Aiken suggests their effect is "to throw doubt upon the validity of the rules themselves. And in that case, there is usually no alternative to a fundamental reconsideration of the whole moral code" (Aiken, 1962, p. 75).

Justification of an entire moral code occurs at the third, ethical level of moral discourse. Aiken terms this process ethical criticism. Even here, at the pinnacle of meta-ethical analysis, the concern remains practical--the resolution of contradictory reasons for implementing a particular moral decision, the removal of moral ambiguity. But, as Aiken has argued, removing our practical doubts requires not only the rational tools of ethical deliberation, analysis, and criticism, but also the uniquely human impulse that motivates the will into action and facilitates our intuitive grasp that events require our moral attention.

The diverse sources surveyed thus far suggest that emotion plays several important roles in morality: early detection system, facilitator of our morally perceptive abilities, and moral motivation and perseverance during the sequence of moral discourse and investigation. The question to be answered, however, is how to resolve such moral dilemmas, especially those involving the protection of human subjects, and to what extent does the framework that Aiken espouses relate to the process of ethical reflection in which those involved in the protection of human subjects engage.

Applying Aiken's Levels of Moral Discourse to Work in Protecting Human Subjects

According to the framework Aiken presents, when designing, reviewing, or conducting research, we begin with our initial, emotional reaction to the protocol. Perhaps the design will involve a placebo-control arm for an anti-infective trial to be conducted at a variety of sites, several of which are located in developing nations. The person designing such a trial may feel an initial gut reaction against such a design. So might the person who is called on to review and approve such a protocol on an IRB or the person selected in the developing nation to actually supervise the conduct of the trial. But why? And, do these initial moral objections from three different persons have the same basis?

To reject, out of hand, such a trial design at this stage of moral consideration would, according to Aiken, be inappropriately hasty. In order to answer the questions posed, we need more information, and we must appraise that information in light of existing rules, moral standards, and regulations. We must escalate our initial emotive reaction to the second level of moral discourse. Perhaps the protocol author consults the International Conference on Harmonisation (ICH) Good Clinical Practice (GCP) guidelines, the set of regulations that usually govern international clinical trials. These guidelines do not explicitly prevent a placebo-controlled arm, but a full reading and understanding of them does suggest to the trial designer several items of moral consideration, namely:

1. Protections for vulnerable subjects (ICH E6, 1996, regulation 1.61)

2. The rights, safety, and welfare of subjects takes priority over the interests of the science (ICH E6, 1996, regulation 2.3)

3. Only qualified individuals should be involved in trial design (ICH E6, 1996, regulation 5.4.1)

Since the trial will likely involve impoverished persons in developing nations, the designer should ensure that additional protections are in place to safeguard these vulnerable persons. Perhaps an independent data monitoring committee should be used to monitor the safety of all subjects, for example. For the sake of argument, let us say that the addition of this safeguard assuages the protocol designer's initial objections to the trial design, and he submits the protocol for approval at the local sites. This situation is analogous to Aiken's claim that rarely do we need to move beyond the second level of moral discourse in order to reach a resolution to our moral ambiguity. In this case, consulting relevant rules and regulations helped guide the protocol designer's inner moral dialogue.

Perhaps one of the protocol reviewers questions why the trial needs a placebo-controlled arm at all. Perhaps he or she believes that despite all of the safeguards the trial designers have added, it is unjust to subject particular volunteers to a placebo when there are other, beneficial alternatives available to participation in the trial. In order to receive approval to conduct the research, perhaps the reviewer mandates that the company provide the most efficacious anti-infective currently licensed to the local population not participating in the study. Such a suggestion would represent Aiken's third level of moral discourse, because the options generated at the second level of discourse still meet with unacceptable consequences and further dialogue is required. With this suggestion, the reviewer is engaging in a creative reconceptualization of the ethical principle of justice. It is not merely enough to ensure a proper balance of benefits and burdens on the population to be selected; benefits to those in society at large should be provided to counterbalance the potential burdens to volunteers who might receive a placebo.

Perhaps, as a matter of science policy, the reviewer strongly believes that such considerations of respect for persons, beneficence, and justice should be built into any protocol that seeks to use human volunteers in developing nations. Changing the Declaration of Helsinki, the Council for International Organizations of Medical Sciences (CIOMS) Ethical Guidelines, or other international codes of research ethics would represent an attempt at achieving Aiken's fourth level of moral discourse. Surely, championing such changes would require additional debate and consideration at many levels, but having the moral courage, stamina, and perseverance to attempt such systematic change on a global level would foster the sort of moral discourse that strives to improve human flourishing, which should be ultimate goal for ethical discourse.

Conclusion

Utilization of Aiken's four levels of moral discourse can assist those called on to review research from an ethical standpoint gain clarity around their own ethical intuitions and work better as a group to reach decisions about ethical issues during IRB meetings. In addition, use of the framework can help instruct those who design and conduct such research in considering the ethical principles of respect for persons, beneficence, and justice from the perspective of the human subjects who will volunteer for such research. Table 1 illustrates the progression through Aiken's four levels of moral discourse as a tool that can add ethical substance to human subject protection considerations that begin with the initial feeling of "Uh, oh," "Hmm," "Ugh," "Yuck," or other indications of moral ambiguity. The second level would feature infusion of relevant rules, regulations, guidelines, and consideration of ethical principles. The third level would include interpretation and analysis of the ethical principles from the Belmont Report and other sources in relation to the research under design, review, or implementation, and perhaps reconceptualization of these principles in a new light to yield resolutions that might, at the fourth level, increase human flourishing and health, which is, after all, the ultimate goal of all research.

Author's Note and Acknowledgement

This manuscript does not represent the views or policies of the NJ Department of Health & Senior Services or the State of New Jersey. All views expressed in the manuscript are the author's own. This manuscript received First Place for Best Paper in the Society of Research Administration International's 2004 Symposium. The author wishes to thank the anonymous reviewers for the 2004 Symposium. The author wishes to thank Mary Faith Marshall, Ph.D., who first introduced the author to the work of Henry David Aiken in 1998, and Joseph Fins, M.D., who suggested the similarities between John Dewey's pragmatism and the work of Aiken. The author also wishes to thank several reviewers for their comments on the manuscript: Lisa Hovey and Erica Rose, J.D.

References

Aiken H.D. (1962). Levels of moral discourse. Reason and Conduct: New Bearings in Moral Philosophy. Westport, Conn.: Greenwood Press.

Callahan, S. (1988). The role of emotion in ethical decisionmaking. Hastings Center Report 8,3, (pp. 9-14).

Callahan S. (1991). In Good Conscience: Reason and Emotion in Moral Decision Making. San Francisco, Calif.: Harper.

Fins, J.J., Bachetta, M.D., Miller, F.G. (1997).Clinical pragmatism: A method of moral problem solving. Kennedy Institute of Ethics Journal 7,2, (pp. 129-145).

Fletcher, J.C., Miller, EG., & Spencer, E.M. (1998). Clinical Ethics: History, Content, and Resources. In: Fletcher, J.C., Lombardo, P.A., Marshall, M.F., & Miller FG. (Eds.) Introduction to Clinical Ethics. (pp. 3-20). Frederick, Maryland: University Publishing Group.

Hayes J.R. (1986). Consultation-liaison psychiatry and clinical ethics: A model for consultation and teaching. General Hospital Psychiatry 8, 6 (pp. 415-418).

International Conference on Harmonisation (ICH). (1996). Guidance for Industry: E6 Good Clinical Practice: Consolidated Guidance. Rockville, Maryland: Food and Drug Administration.

Koski, G. (2003). Changing the paradigm: New directions in federal oversight of human research. Journal of Pediatric Gastroenterology and Nutrition 37, Supplement, (pp. S2-SS6).

National Commission for the Protection of Human Subjects in Biomedical and Behavioral Research. (1979). The Belmont report: Ethical principles and guidelines for the protection of human subjects. Washington, DC: Department of Health, Education, and Welfare.

Purtilo R.B. (1999). Ethical Dimensions in the Health Professions. Philadelphia: W.B. Saunders Company.

Sherman N. (1990). The Place of Emotions in Kantian Morality. In: Flanagan, O. & Rorty, A.O. (Eds.) Identity, Character, and Morality: Essays in Moral Psychology. (pp. 149-170). Cambridge, Mass.: MIT Press.

Wocial L.D. (1996). Achieving collaboration in ethical decision making: Strategies for nurses in clinical practice. Dimensions of Critical Care Nursing 15, 3, (pp. 150-158).

David Perlman, Ph.D.

Director, Human Research Ethics Program

NJ Department of Health & Senior Services

And

Adjunct Assistant Professor

University of Pennsylvania School of Nursing

PO Box 360

Trenton, NJ 08625

Phone: [phone]

Email: [e-mail]

Table 1. The Four Levels of Moral Discourse

Level 4        Level 1
(Emotion)      (Emotion)

[up arrow]      [down arrow]

Level 3        Level 2
(Reason)       (Reason)

       [left arrow]