Supererogation in clinical research
- First Online:
- Cite this article as:
- Barnbaum, D.R. Med Health Care and Philos (2008) 11: 343. doi:10.1007/s11019-008-9124-4
‘Supererogation’ is the notion of going beyond the call of duty. The concept of supererogation has received scrutiny in ethical theory, as well as clinical bioethics. Yet, there has been little attention paid to supererogation in research ethics. Supererogation is examined in this paper from three perspectives: (1) a summary of two analyses of ‘supererogation’ in moral theory, as well as an examination as to whether acts of supererogation exist; (2) a discussion of supererogation in clinical practice, including arguments that both physicians and patients can practice acts of supererogation; (3) a discussion as to why researchers, qua researchers, are not routinely recognized to perform acts of supererogation, while at the same time the very nature of research subject participation involves supererogation. The article concludes by considering three examples of supererogation on the part of researchers, with a plea that researchers’ supererogatory actions be recognized as such.
KeywordsEthical theory Physicians Research ethics Research participants Supererogation
Supererogation in research
The concept of ‘supererogation’ is familiar to ethical theorists, somewhat recognizable to those in clinical practice, and woefully foreign to researchers. What is supererogation, and why does it remain obscure in research ethics?
An answer to the above question unfolds in three steps. The first requires an examination of ‘supererogation’. Competing but similar notions of this concept are presented. One of the most controversial questions in ethical theory about supererogation is whether it exists at all. Perhaps the concept is meaningful (there is a connotation), but perhaps there are no actual acts of supererogation (there is no denotation). Both the concept of supererogation, and the debate as to whether acts of supererogation exist, are considered.
The second step is to examine supererogation in clinical practice. Just as there are debates in ethical theory about what ‘supererogation’ is, and whether acts of supererogation exist, a debate in bioethics exists about the status of supererogation in clinical practice. In particular, there is significant debate as to whether physicians can, or do, practice acts of supererogation. Both sides of this debate are presented. Patients, and not just healthcare providers, may perform acts of supererogation, so an examination of those acts is also warranted.
The third step is to ask why no discussion similar to the above exists in research ethics: a discussion about healthcare practitioners and supererogation exists, but a discussion about researchers and supererogation is yet to begin. It is an odd anomaly, since the ethical concepts employed in clinical practice and in research are often similar. For example, debates exist about what ‘beneficence’ means in both clinical practice and in research, the scope and limits of beneficence in both treatment and research, and whether beneficence can be overridden by other principles in both treatment and research. Similar debates exist across both clinical ethics and research ethics about a multitude of concepts and practices: informed consent; autonomy; protecting the interests of those with compromised autonomy; justice. While the answers to questions in clinical ethics and research ethics may not be the same, there are significant commonalities, a shared moral language. And yet, while questions about supererogation have arisen about health care practitioners, supererogation has been absent from debates about the moral responsibilities of clinical researchers.
The anomaly does not persist with respect to patients and research participants, however. Both are recognized to perform acts of supererogation by virtue of their respective roles. Four suggestions are offered to account for the fact that patients and research participants, as well as clinicians, routinely perform acts of supererogation, although people have been slow to recognize supererogation on the part of researchers. The discussion concludes by suggesting three places in which supererogation in clinical research might exist, and a plea to recognize these contributions as supererogatory.
The concept of ‘supererogation’
‘Supererogation’ is taken from a Latin word which means ‘to overspend’ (Mellema 1991, 14). In a contemporary context acts of supererogation are those that are morally exceptional, those which go beyond the call of duty. Discussion of supererogation in ethical theory was revived in 1958 by J.O. Urmson in his article “Saints and Heroes” (Urmson 1958). Although he did not use the term ‘supererogation’ in his article, Urmson called for examination of a category of moral evaluation beyond rightness, wrongness, and obligation. Surely there should be space for the moral saint, or the moral hero, one who does the exceptional, either “by control of contrary inclination and interest” (or) by control of natural fear or without effort.’ (Urmson 1958, 201).
Contemporary analyses of supererogation attempt to carve out that space. Gregory Mellema analyzes an act of supererogation as one which (a) does not fulfill any duty or obligation, (b) is morally praiseworthy, and (c) if omitted, is not morally blameworthy (Mellema 1991, 13–14). Mellema adds that a desirable addition to the above analysis would be a “continuity requirement’: acts of supererogation fulfill more of the same type of value as the action that duty would require an agent to perform (Mellema 1991, 22–29). In other words, acts of supererogation are those that are on the same scale of value as other actions, only higher.
A second analysis of ‘supererogation’ is presented by David Heyd. Heyd contends that an act of supererogation (a) is neither obligatory nor forbidden, (b) is an act whose omission is not wrong, nor does its omission deserve sanction or criticism, (c) is an act whose performance is good, by virtue of its intended consequences and its intrinsic value, and (d) is performed voluntarily for the sake of someone (else’s) good (Heyd 1982). A.C. McKay adds an additional condition to those Heyd offers: “the act should have a potential cost to the agent. By undertaking the act, the agent risks losing something which he might reasonably be regarded as being entitled to, on some scale of values that includes his own interests.” (McKay 2002).
While supererogation is most often cited as a concept in normative ethics—the ethics of the rightness and wrongness of actions—the concept is occasionally cited in virtue ethics as well. For example, Thomasma and Kushner propose that supererogation “is the intensification of a virtue beyond that called for in any duty one either has by reason of being human or by reason of a special public or private commitment one makes.” (Thomasma and Kushner 1995). Their discussion considers whether the virtue of compassion is supererogatory for physicians, arguing that while compassion may not be necessary to cure, it is nonetheless essential to heal, which is a higher calling in medicine than merely to cure. But discussions of supererogation in virtue ethics are comparatively rare. Thus, the focus here will remain on Mellema’s, Heyd’s, and McKay’s analyses of supererogation as a property of actions, not of virtues.
Mellema’s first three conditions are about actions and obligations, whereas his continuity requirement is about consequences and outcomes. Heyd’s first two requirements are about obligation, and his third requirement is about intentions. It is notable that Mellema does not address intentions at all. Finally, Heyd’s final condition is about beneficence—a familiar concept in bioethics. McKay’s addition to Heyd’s analysis incorporates claims about agents’ intentions. Mellema does not address the concept of beneficence in his analysis, nor intention. What each analysis has in common is the ordinary-language sense of an action that is not morally required, but is morally extraordinary.
While the connotation of an act of supererogation—one which is exceptional, beyond the call of duty, not morally obligatory but nonetheless permissible and praiseworthy—is clear, the denotation of such acts remains a vexed question. A significant difficulty locating acts of supererogation is that two prominent moral theories leave no room for acts of supererogation.
Kantian deontology claims that an act is morally permissible if and only if it follows from the sole moral rule, the categorical imperative: act only on that maxim that can at the same time be willed to be a universal law. If an agent’s action is not consistent with the categorical imperative, then that action is morally wrong. If the action is consistent with the categorical imperative, then the action is morally right, morally permissible. But there is no room according to Kantian deontology for actions that are permissible, but optional, actions that follow from the categorical imperative, but if an agent fails to perform them, the agent’s failure is not blameworthy. As stated by McKay, “A strict interpretation of Kant’s categorical imperative, “act only on that maxim which (one) would at the same time will as a universal law”, might also seem to imply that, as duty is all-pervasive, the concept of exceeding it is empty.” (McKay 2002, 70). Thus, Kantian deontology leaves no room for the supererogatory.
According to utilitarianism, an action is morally right if and only if it maximizes the best outcome for all parties concerned. Utility maximization may take the form of promoting happiness, pleasure, or some other intrinsic good. Any failure to maximize the greatest good for all parties concerned is a failure to perform the right action according to utilitarianism. Thus, there are no actions that are morally permissible and which promote even greater good than some action which is already morally permissible on the same scale of value. If there was some alternative action (call it action2) to the action that maximized utility (action1) which promoted even greater good than action1, then action2 would in fact be the action that maximized utility, thus being the action that was morally required, not action1. In fact, it makes no sense at all to say “Action1 results in the greatest utility, so it is morally right to do, but some other action, action2, results in even greater utility, making action2 supererogatory.” Thus, according to utilitarianism, there are no acts of supererogation.
Some prominent moral normative theories are troubled by the notion of supererogation. For those who examine morality from a bottom-up perspective, however, looking at cases first rather than theories, supererogation may be rare but not too hard to find. Fathers trek out in the cold and snow to cover a several-mile walk in an attempt to save stranded family members; strangers dive into a subway well and huddle between the tracks to save the life of an incapacitated student as a train approaches. Perhaps it is so much the worse for Kantian deontology or utilitarianism that these moral theories have difficulty accommodating acts of supererogation.
So the debate about supererogation proceeds: on the one hand, some moral theories have a hard time accommodating such actions. On the other hand, intuitively, we believe that acts of supererogation exist. Perhaps we just know supererogation when we see it. Bioethicists believe that they have been seeing it in clinical practice for a long time.
Supererogation in clinical practice
There are two places to look for supererogation in clinical practice: the actions of health care practitioners, and the actions of patients. If acts of supererogation exist at all, it would appear that both health care practitioners and patients are in a position to perform those actions.
McKay holds that “the duties demanded of the individual physician by the standards set for the profession by itself are supererogatory and hence, in Heyd’s terms, deserving of praise and respect.” (McKay 2002, 72). McKay’s view is that no one is obligated to become a physician, yet in becoming a physician an individual commits to duties that both are open-ended and necessitate a relinquishing of self-interest. For these reasons, being a physician is what Aquinas would have conceived of as one of the “aristocratic vocations”, one which by its very nature requires an agent to go beyond the call of duty (Stanford Online Encyclopedia of Philosophy).
… beneficence prescribes an obligation to act in a certain way, whereas altruism prescribes no such obligation but is instead optional and supererogatory, beyond the call of duty…. But physicians have a special relationship with their patients, and this relationship does create specific duties, such that in their routine clinical practice, physicians are not altruistic. (Glannon and Ross, 68)
Glannon and Ross move away from the notion of physician as an aristocratic vocation, one which, by its very nature entails saintly or heroic actions. Their position echoes M.H. Kottow’s, who holds that “medicine is duty-bound to benefit patients and that supererogation/altruism must per definition remain outside and beyond any role-description of the profession.” (Kottow 1990). But Glannon and Ross’s observation is telling: given that a special relationship exists between physicians and their patients, this relationship requires moral actions that might otherwise be seen as exceptional. Physicians’ actions in treating their patients are not supererogatory, they are morally obligatory. But what about cases in which physicians act beyond the “specific duties” that Glannon and Ross cite, or in which physicians treat patients towards whom they have no special relationship, or perhaps, no relationship at all? Such actions would appear to be supererogatory. For example, physicians who volunteer for organizations such as Doctors Without Borders are acting in a supererogatory fashion, precisely because they are serving individuals with whom they have no “special relationship”, and in many cases, serving those with whom they previously had no relationship at all.
While Glannon and Ross do not believe that physicians perform acts of supererogation, they do believe that patients do, such as when patients allow themselves to be used by medical students for teaching purposes. Glannon and Ross see this action as one of heroic (but not saintly) beneficence, as no patient is obligated to be used for teaching purposes. Persons who are living organ donors for strangers certainly perform acts of supererogation, ones of a more “saintly” variety in Glannon and Ross’s estimation. These donations capture Mellema’s requirement that the act fulfills no duty or obligation, as well as Heyd’s requirements that the performance is good by virtue of its intended consequences and intrinsic value, and that the action is done voluntarily for someone else’s good. These donations are perceived to be so morally exceptional that the donors are often suspected of having ulterior motives. Such actions clearly exceed the moral obligations that strangers typically have towards one another.
In summary, when supererogation occurs in the clinical setting, it does so either by virtue of actions that exceed those typically required by those in a given relationship, or by virtue of actions that exceed the beneficence typically required by those performing the actions. Doctors Without Borders perform acts of supererogation, because no previous relationship existed; living organ donations to strangers both exceeds the moral obligations inherent in the relationship between strangers, and exceeds the level of beneficence that characterizes most moral obligations. Both clinicians and patients perform acts of supererogation. How do such lessons play out in the research setting?
Supererogation in the research setting
A conversation exists about supererogation in the clinic, but the conversation has not yet started in research ethics. Some may argue that research itself is supererogatory—that researchers are engaging in an aristocratic vocation. Daniel Callahan disagrees with this assessment (Callahan 2003). Callahan holds that performing research is an “imperfect duty”. Imperfect duties, as presented by Kant, are those moral obligations that one has, but one is not required to fulfill every time that an opportunity presents itself. For example, Kant held that giving to charity was an imperfect duty: it is morally right to give to charity, but it is not the case that every time a person has a spare dollar in his pocket that he is obligated to give it away. In Callahan’s view, research is an imperfect duty, it is not continuously obligatory. Furthermore, Callahan does not believe that the failure to perform research is blameworthy—in his words, the “research imperative” is misguided at best, and dangerous at worst. Two of Mellema’s three necessary conditions for supererogation (research is not obligatory; the failure to perform research is not blameworthy) are fulfilled according to Callahan’s account. But Callahan also believes that most research is not praiseworthy, thus eliminating research as supererogatory on Mellema’s account. Just as Glannon and Ross, as well as Downie, argue that physician is not an aristocratic vocation, so too Callahan argues that researcher is not aristocratic.
When it comes to researchers, there is nothing in the literature that argues their actions, qua researchers, are supererogatory. Not so with respect to research subjects. The most obvious example of research participation as supererogatory is non-therapeutic research. Research subjects in phase I and phase II clinical trials in the United States, in which the prospect of direct medical benefit is small, and in some cases non-existent, are clearly performing actions that exceed the typical requirements of beneficence. While direct therapeutic benefit may be low, or non-existent in such studies, other “benefits” may accrue to subjects in such trials, such as free medical care or cash payments. These “benefits” are not recognized as benefits by regulatory agencies such as the FDA or OHRP, and as such are not calculated as part of a risk/benefit assessment when determining if research protocols are ethically permissible. Some research subjects may nonetheless view free medical care or cash payments as beneficial. 1 However, even when these “benefits” are made available, the therapeutic benefits of a phase I or phase II clinical trial are limited, even as the therapeutic risks remain significant. As such, participation in phase I or phase II clinical trials is often viewed as supererogatory. This point is illustrated by the awe inspired by researchers who serve as their own research subjects, certainly without the “benefits” of cash payments or free medical care. Callahan calls researchers who served as their own subjects in the effort to cure yellow fever examples of “research heroism” (Callahan 2003, 136). Lawrence K. Altman’s Who Goes First? The Story of Self-Experimentation in Medicine serves to praise those who bravely served as subjects in their own research. Even the Nuremberg Code’s famous admonition that researchers cannot use subjects in experiments that prima facie are believed to kill or seriously harm patients, unless the researchers serve as subjects themselves (Nuremberg Code 1948), hammers home the point: research subjects, especially those who serve when the prospect of direct therapeutic benefit is low, are surely acting in a supererogatory fashion.
It can be argued that all research subjects, and not merely those who have little or no prospect of direct therapeutic benefit, act in a supererogatory fashion. Even when there is a likely prospect of direct therapeutic benefit, this benefit is not guaranteed. If it were guaranteed, then there would be no need for research in the first place. So if there is no guarantee of direct therapeutic benefit, what is the basis of a positive risk/benefit ratio? A positive risk/benefit ratio exists when benefits might accrue to future cohorts of the research subjects. Such benefits have been referred to as “aspirational benefits”—benefits that research subjects might not enjoy, but that future cohorts will (King 2000). Studies might, but might not, directly benefit research subjects. But even when subjects themselves do not receive direct therapeutic benefits, they may be altruistically benefiting future cohorts, altruism which Glannon and Ross cite as supererogatory. Thus, research subjects routinely perform acts of supererogation, simply by virtue of being research subjects, since their actions demonstrate exceeding beneficence that can be characterized as altruistic.
It may be argued that in some cases subjects have affinities for future sufferers of the same condition that they have, hoping to prevent others from experiencing the same pain and hardship that they are experiencing. Barry Brown makes much of this cohort relationship, arguing that it is the basis for the morally permissible use of some research subjects who are not competent to consent to participation (Brown 2006). Brown argues that the surrogate decision making for these incompetent subjects on the basis of a substituted judgment is warranted. Given that subjects have affinity for their cohorts—even if those subjects are incompetent—that this affinity serves as a basis for the claim that if the subjects were able to offer informed consent, then they would. If Brown is correct, then research that has mere aspirational benefits nonetheless directly benefits each subject, since that subject has an interest in benefiting his cohorts.
But what is the basis for these subjects’ affinity for their future cohorts? Brown’s claim is that there is not only a present community of Alzheimer’s patients, or cancer patients, but that this current community shares interests with future patients, and even shares interests with those persons who might not get these diseases in the future, but would have if not for members of this community. Taken to its logical and absurd conclusion, everyone could share community with everyone else. Everyone alive today is a potential carrier of polio, all owing a debt to those who first developed and tested a polio vaccine, but every one of us should hardly be considered cohorts of those who first tried the vaccine. In such a case, the words “community” or “cohort” fail to have any meaning.
In summary, research subjects routinely perform acts of supererogation, merely by being research subjects. Research subjects may not therapeutically benefit from serving as research subjects, so their actions may exceed the typical bounds of beneficence. Even in cases in which participants obtain non-therapeutic benefits, such as cash payments or free medical care, those “benefits” are outweighed by the risk of medical harm in phase I or phase II studies. Furthermore, it can be argued that research subjects do not always share community with their future cohorts, thus their participation as research subjects exceeds what is typically morally required of such a relationship. The same lessons from the supererogatory actions of physicians and patients apply here: research subjects perform actions that either exceed typical duties of beneficence, or exceed typical duties that emerge by virtue of relationships with others. Thus, their actions, qua research subjects, are supererogatory. For these reasons, we should never hesitate to show research subjects the respect that they deserve, in virtue of their supererogatory acts.
The search for generalizable knowledge requires painstaking adherence to research protocols. Every protocol deviation should be reported, in part because deviations could jeopardize the efforts of the entire research team. As such, any attempt to do less, or more, than a protocol demands, may have a negative impact on everyone’s efforts. Thus, there is no room for going beyond the call of duty in research.
As argued above, research protocols may not provide therapeutic benefit to participants. The claim that subjects participate in research to benefit their cohorts, and thus their participation is not entirely altruistic, does not hold up. Additionally, protocol deviations are reported with the assumption that these deviations harm subjects, either directly by putting an individual subject at greater risk, or indirectly by jeopardizing the quest for generalizable knowledge. These assumptions reinforce the belief that being a research subject is a burden, not a benefit, even in therapeutic research that may directly benefit research subjects. Thus, research subjects, merely by serving as research subjects, perform acts of supererogation. No such special designation exists for researchers.
Today’s research often involves tremendous financial investment and collaboration across multiple sites. For these reasons, it is enormously difficult for researchers to strike out on their own, doing what was at one time exceptional. Altman’s daring examples of researchers who use themselves as research subjects are, in large part, creatures of the past. Research is very much a collaborative effort. Because of the collaborative nature of research, researchers are hard-pressed to do what is exceptional and go beyond the call of duty.
Finally, as all researchers know, the regulations and paperwork required to perform research are overwhelming. Many researchers are stretched to their limits to do what they are required to do, let alone to do anything above and beyond what is required.
All of these reasons reinforce the notion that while acts of supererogation may be open to research subjects, researchers should be content with doing their moral duty, and not more than that.
Some may believe, however, that supererogation is available to researchers, just as it may be for clinicians. It was observed earlier that neither Kantian deontology nor utilitarianism can accommodate acts of supererogation. But clinical research is not guided exclusively by either of these theories. Thus, there may be hope yet to find acts of supererogation available to researchers.
Three possible avenues for supererogation on the part of researchers exist. One possibility is the notion of informed consent as a process, and not as a one-time event. Rather than thinking of “consenting” as an event that happens, and then is completed, researchers can think of consent as a process that transpires across the entire duration of a subject’s participation. Every interaction with a research participant can thus be an opportunity to reiterate the goals, process, risks, benefits, and reaffirm the voluntary participation of the participant. So conceived, informed consent transcends that which is obligatory on the part of researchers, is praiseworthy, and is done for the sake of someone else’s good.
A second example of supererogation is the action of performing research that might put the researcher in harm’s way, or act in ways that benefit individuals with whom the researcher has no prior relationship. For example, researchers who travel to distant continents to research highly contagious and fatal diseases in persons with whom they had no prior connection are acting in a supererogatory fashion. Mellema’s analysis of ‘supererogatory’ does not support this claim because, as was pointed out earlier, Mellema’s analysis does not incorporate intention nor beneficence on the part of the agent performing the act of supererogation. Rather, Mellema’s analysis focuses solely on properties of the supererogatory action, and not the agent who performs the action. Heyd’s and McKay’s analyses are more helpful in this regard. Heyd’s analysis includes the condition that the agent acts voluntarily for the sake of someone else’s good. McKay supplements Heyd with the expectation that supererogatory actions are contrary to the agent’s own interests. Thus, both Heyd and McKay help to identify the researcher who for example, travels to Africa to research a cure for Ebola, as one who is acting in a supererogatory manner. Glannon and Ross argue that physicians have special duties to their patients, in virtue of their special relationship, although the duties that arise in virtue of this special relationship do not distinguish physicians’ actions as supererogatory. This observation points us in the direction of considering physicians who do not have a prior “special relationship” to their patients as performing acts of supererogation, such as physicians who volunteer for Doctors Without Borders. Similarly, researchers who volunteer to serve those persons with whom they had no prior “special relationship” are acting in a supererogatory fashion.
Third, researchers who give of their time to further other people’s research also perform acts of supererogation. Researchers who serve on IRBs, DSMBs, or grant review committees contribute to research, albeit not their own. Many conflict-of-interest guidelines carefully and elaborately guarantee that the researchers serving on these committees have no interests in furthering the research they are reviewing. Again, in keeping with Heyd’s and McKay’s analyses, these actions are done voluntarily for the sake of someone else’s good, and may be contrary to the agent’s own interests. While perhaps not as “saintly” as traveling to Africa to perform Ebola research, anyone who has served long hours on an IRB, DSMB, or grant review committee will agree that their commitment was undertaken for the sake of someone else, and that the investment of time may well have been contrary to their own interests.
In each of these three cases, the same lessons about supererogation emerge. Researchers act in a supererogatory manner when their actions exceed a relationship, or when they exceed beneficence, be it incorporating informed consent that is a process and not merely a one-time event, working for strangers in difficult conditions, or undertaking committee work in which they have no personal investment. Researchers should be commended for these supererogatory actions, recognized for performing actions that go beyond the call of duty.
The shared moral language that guides clinical practice and clinical research includes the concept of supererogation. Physicians, through exceptional acts of beneficence that exceed what is typically to be expected given the nature of their relationship with the patients, have the opportunity to perform acts of supererogation. The debate as to whether physicians engage in an “aristocratic” vocation may be unsettled, but no one can dispute that the very nature of clinical practice involves working for the benefit of others with whom the physician had no previously existing relationship, in many cases at great personal cost. Patients can also act to benefit others with whom they had no previous relationship, whether in their availability to help teach the next generation of physicians, or to generously donate organs for the benefit of fellow patients. Thus, patients too have the opportunity to perform acts of supererogation. The same chance to act beneficently for the sake of others emerges for research participants. In clinical research—in particular non-therapeutic clinical research—research participants routinely are recognized for acting voluntarily for the sake of someone else’s good, at potentially great cost to themselves. But while there may initially seem to be structural barriers preventing researchers from practicing acts of supererogation, closer analysis demonstrates three examples in which researchers have the opportunity to go beyond the call of duty. When exceeding the typical practices associated with informed consent, performing research that could come at great personal cost that benefits those with whom the researcher previously had no “special relationship” or helping to further others’ research agendas, researchers perform acts of supererogation. Such praiseworthy actions should be accorded the moral respect that they deserve.
I wish to thank an anonymous reviewer for recommending that I make this observation.
Earlier versions of this paper were presented at two conferences sponsored by Clinical Research Management, one at the Walter Reed Army Institute for Research, Silver Springs, Maryland, in October 2006 and the other in Cleveland, Ohio, in December 2006. Additionally, I wish to thank two anonymous reviewers at Medicine, Healthcare and Philosophy for their helpful comments and suggestions.