Toward a General Theory of Research Ethics

Two seemingly distinct principles have dominated the ethics of human subjects research. On the one hand, based on the importance of personal autonomy, we require the informed consent of subjects who have decisional capacity, and provide special protections to those who lack such capacity. Separately, however, we also impose a variety of substantive restrictions on the conduct of studies.

But do we really need two principles? Franklin Miller and Alan Wertheimer have produced a rich, elegantly argued essay suggesting that much of research ethics, as it relates to studies intended for adults with decisional capacity, can be derived from a single intriguing concept. They name it "group soft paternalism."

Their line of reasoning proceeds as follows: First, because we cannot always accurately determine when someone has decisional capacity, people who lack sufficient capacity are regularly enrolled in such studies. To protect them from harm, we allow the studies to be conducted only if they adhere to various substantive restrictions, including specified risk-benefit criteria. These rules infringe on the autonomy of fully competent subjects. They prevent them from enrolling in studies that might otherwise have been conducted if there weren't any risk-benefit restrictions. The name "group soft paternalism" refers to the general policy of protecting from harm the members of a group that have impaired judgment, but for practical reasons applying that policy to all members of the group.

Much of Miller and Wertheimer's discussion is devoted to an intricate and convincing demonstration of why group soft paternalism-which, as they forthrightly acknowledge, some people would claim is not truly paternalism-does not unethically infringe on anyone's autonomy. But one of their further claims-that this construct provides an explanation for the existing research rules relating to risks and benefits-seems less convincing.

Imagine that the existing rules had been specifically crafted to deal with the authors' hypothesized scenario that many people who lack decision-making capacity are regularly enrolled in studies designed for competent adults. If that were the case, then those rules have a huge ethical flaw. Why? Because the risk-benefit rules that Miller and Wertheimer are trying to justify don't require that, from a subject's viewpoint, the benefits outweigh the risks. Far from it. They merely require that the benefits to the subject, combined with the benefits to society from the knowledge to be gained, must be reasonable relative to the risks. Depending on the importance of that knowledge, the risks to a subject can substantially outweigh the benefits to that subject. It is hard to believe that the architects of the federal regulations would have intentionally created a system that accepts regularly exposing incapacitated subjects to such harms.

And it is not as if it would be impossible to design risk-benefit rules that adequately protect those "wrongly enrolled" incapacitated subjects. We have created exactly such rules for studies where we do expect enrollment of subjects who won't all be able to give meaningful consent. Perhaps the best example is the rules for studies involving minors. Indeed, Miller and Wertheimer's concept of group soft paternalism serves remarkably well as a justification for those rules. Not all minors lack capacity, and some might want to exercise their autonomy to enroll in studies where risks might substantially outweigh benefits. But because it might not be feasible or cost-effective to specifically identify the minors with capacity, we instead impose strict risk-benefit criteria that limit the autonomy of all minors who enroll in research studies.

In contrast, it seems unlikely that the rules for studies involving competent adults contemplate the enrollment of many incompetent subjects or are justified by group soft paternalism. The justification likely requires looking to true-hard-paternalism, and perhaps even to reasons based on societal needs, and not merely on the best interests of the subjects. And in evaluating those justifications, we shouldn't be too quick to assume that the ethical rules for research are all that different from those we apply to medical practice. While patients have the right to refuse treatment, there are very substantial restrictions on their right to demand a...