March 19, 1999

By Steven E. Hyman, M.D.

Procedures routinely used to safeguard the rights and well-being of human subjects in research on mental illness have been criticized during the past year in the press, in Congress, and, by implication, in restrictions on psychiatric research recommended by the National Bioethics Advisory Commission (NBAC). The research community and some patient advocacy groups have responded that serious abuses are rare and that the proposed restrictions would impede needed clinical research. These rejoinders have not turned the tide; rather, they have been interpreted as defensive and as unresponsive to the concerns of Congress and of the Bioethics Commission.

I believe that our core values provide us with a far better response: Respect for and protection of human subjects, our partners in research, are fundamental to all psychiatric investigation. We are well aware that mental illnesses are destructive, disabling, and, all too often, lethal. It is simply unacceptable to lock our patients and their families into the status quo. Thus, working with patient volunteers to discover causes and develop better treatments for schizophrenia, autism, manic-depressive illness, and many other severe mental disorders is the principal challenge confronting our field.

Given this challenge, my goal is to establish procedures for protecting human subjects in research funded and conducted by NIMH that will serve as a model for all medical research that uses human subjects. Staff of NIMH, along with many of our investigators, recognize that studies that involve vulnerable populations should trigger a higher level of concern by the local Institutional Review Boards (IRBs) that oversee research involving human subjects. NIMH has been proactive in this arena. Over the past 14 months, NIMH has sponsored two workshops to address issues involving research with vulnerable populations, and we are planning a third. The workshops held to date have been productive and illuminating and are having an impact on our policies.

Our first workshop, held in December 1997 in conjunction with other institutes in the National Institutes of Health (NIH), focused on individuals with questionable capacity to consent. Presenters argued that increasing levels of decisional impairment and study risk should prompt IRBs to exert increasing levels of scrutiny and to employ additional safeguards on a sliding scale. Additional protections (for example, involvement of family surrogates and/or independent monitoring of the consent process) may be highly advisable in certain situations. But the presumption that all individuals who have a mental disorder or cognitive deficit are incapable of understanding research or exhibiting altruism is both inaccurate and disrespectful.

Workshop panelists pointed out the need to find better ways to assess not only potential participants' cognitive grasp of a study, but also their appreciation of precisely how a study's risks and benefits apply to them. Panelists also emphasized that consent is not synonymous with the signing of a document, but rather must be an ongoing process that should involve education of the potential research participant and, where appropriate, of family members. Panelists also noted that researchers dealing with vulnerable individuals must make a conscientious effort to distinguish research from clinical care.

IRBs that oversee a substantial number of protocols involving participants with mental illness would be advised to have one or more members from the community with experience in the area of mental illness. While providing IRBs with such additional points to consider in dealing with the potentially decisionally impaired, we at NIMH recognize that some IRBs may be overworked and would benefit from innovative forms of technical assistance and "backup." We want to work with NIH to strengthen our IRB system, which is the cornerstone of human studies in the United States.

A second NIMH-sponsored workshop, in December 1998, addressed issues involved in studies that employ medication discontinuation and "challenge" designs. In the former, a research volunteer being treated for a mental illness is taken off medications to permit clinical investigators to examine the subject's functioning in an unmedicated state; to determine the need for continued medications; or to assess the potential side effects of a medication. In "challenge" designs, research volunteers are presented with a pharmacologic or psychological challenge to induce symptoms. The intent of provoking symptoms may be to study the underlying basis of the disorder, to develop or refine a diagnostic test, or to evaluate a novel treatment. In almost all cases approved by IRBs, the symptoms produced by challenges are mild and transient.

Neither medication discontinuation nor challenge designs are unique to psychiatric research. Research of this type was pivotal in developing the cardiac stress test, for example, and the glucose tolerance test. Nevertheless, medication discontinuations or challenge designs should never be used in the absence of appropriate scientific justification and safety precautions. When such designs are employed, in pilot and definitive studies alike, the question under investigation should be well formulated, important, and answerable. Emergency treatment should be readily available during the study in the event that a participant becomes seriously distressed; following the study, subjects should be debriefed about their experience, and there should be follow-up calls or clinic visits to ascertain that no lasting ill effects have occurred.

To ensure that such designs are employed only when appropriate and to maximize the safety and welfare of subjects, NIMH is considering various methods of consultation with independent experts, including not only investigators, but also bioethicists and public representatives, for research designs that are considered high risk.

A third workshop, which NIMH now is planning-and which necessarily will involve the Food and Drug Administration-would examine the proper role for washouts and placebos in treatment trials for conditions for which some effective treatments exist already.

The remarkable progress that we have made to date in understanding and treating mental illnesses would not have been possible without the participation in research of volunteers with mental disorders. Future progress will not be realized without a healthy clinical research effort that will continue to involve participation of people with mental illnesses. Yet even as research on treatments evolves and becomes more effective, our society's attentiveness to the well-being of research volunteers is growing. NIMH and all psychiatric investigators must redouble our efforts to maintain the trust that human subjects place in the excellence and safety of the research enterprise.