As part of a long-range investigation of the social and physiological effects of heroin, Jim Sanders, an anthropologist affiliated with a medical school, set up a study of pregnant women and their infants. The study design involved statistical comparisons of the course of pregnancy, birth, and infant development in 30 addicts and 30 nonaddicts; data were to be gathered from hospitals and public health clinics in three inner city neighborhoods.
The medical school Institutional Review Board (IRB) approved the study, with the requirement that the women sign detailed consent forms indicating that they understood the purpose of the study and agreed to take part. It was agreed that the clinic nurses would read the consent forms to the women and ask them to sign.
Two months into the study, Sanders, visiting the research sites, discovered that few nurses were using the consent forms; instead, informal undocumented consent was being obtained.
Sanders was in a quandary. Institutional Review Boards are designed to protect those studied from harm and, also, to make sure that people do not unknowingly or unwillingly participate in research that might put them at risk. The medical school IRB, accustomed to biomedical experiments posing the risk of physical harm to subjects, routinely required written consent from those studied. In this case, however, the research posed no possibility of physical harm; instead, the only risk to those studied came from the signed consent forms: these would document that addicted women were engaging in illegal behavior.
Should Sanders go back to the Institutional Review Board and attempt to educate them about social research? This might obstruct his project or endanger his research subjects. Should he leave well enough alone, let the IRB think he was following their procedures while allowing the nurses to continue obtaining verbal undocumented consent? In deceiving--or at least, not enlightening--the IRB, he would protect his research subjects. What else might he do to protect his subjects and his study without condoning deception?
Thomas Weaver, University of Arizona, previous chair of Human Subjects Committee (IRB):
Sanders makes several unwarranted assumptions about the risk the study population faces and about the goals and potential actions of the Institutional Review Board. The description of the case states that the only risk faced by the subjects is from the signed consent form. This, of course, is not true. The signed consent form should spell out the potential risks to the subject and, if properly written, should include the potential risks from divulging names or other information received from the subjects. The form or an accompanying document should detail the measures being taken by the investigator to protect the confidentiality of the records and forms, who will have access to them and what will be done with the data once the project has been completed. The subjects should be told about plans to publish the results or whether papers will be given on the material. The subjects should also receive assurance that their names will not be used, and that procedures will be taken to assure that their identity cannot be traced through the consent form or other identifiable characteristics of the population.
The second assumption made by Sanders is that the Institutional Review Board is a regulatory or policing agency of the U.S. government. The fact is that the IRB functions to protect its parent institution and as such is interested in assisting the researcher to do the best research possible while protecting the rights of the subjects of that research. IRBs remind researchers of the necessity to do what they should want to do anyway if they follow their own profession's code of ethics, perhaps suggesting some means to do it better, suggestions which are garnered from having read many other proposals. It seems that Sanders's attitude is one of protecting his subjects from the IRB. Members of the Institutional Review Board are also researchers or persons who have had experience in dealing with similar problems and are most acutely aware of the necessity for protecting the confidentiality of records and the identity of research subjects. The records of the IRB are also confidential, with no one other than staff and members having access.
So, in answer to the questions posed, the IRB has had prior experience with similar cases and is undoubtedly aware of the problems involved. However, there may exist institutions where there is little social science research (Sanders may be in one of these) and so the IRB may have to be enlightened about the special problems involved. Usually the IRB is accustomed to dealing with the special problems endemic to different fields and is open and teachable.
From what has been said above, it should be clear that subjects will not be placed at additional risk by virtue of a discussion with the IRB. In fact, a frank discussion with the Board is called for since the procedures submitted by the investigator and approved by the IRB are not being followed; the investigator is in violation of federal rules and regulations, placing his research and the institution in jeopardy. The rules and regulations under which the IRB and the investigator operate have the force of law and include penalties for their violation.
Other actions that might be taken by Sanders include filing an amended consent form which calls for verbal rather than written consent; here a description must be provided of what each subject will be told, by whom, and signed by the person doing the explanation and, perhaps, a witness. Another possible action is retraining the nurses to handle the consent forms and supervising them to make certain they obtain consent in the approved manner. The IRB may suggest one of these actions when they are made aware of the situation. (Our institution requires periodic reports from investigators to ascertain that approved procedures are being followed. An on-site inspection may be made periodically to assure that appropriate procedures for record keeping are being followed, including the signing of consent forms.)
Whatever Sanders does, he should avoid deception of any kind.
Lucy M. Cohen, Catholic University:
Four questions have been raised about this case. I would like to address, first, the question of the nurses who were the contact persons with the women in the study. In my discussion, I am keeping in mind the fact that Dr. Sanders is two months into the study and that he proposes a "long-range investigation."
1. The question of decisions about "consent of human subjects" within organizations such as hospitals and clinics should not be conceptualized as an issue that involves only a researcher and the formal group established to approve proposals and examine procedures. Research is a process and decisions about how to obtain consent often involve many groups of people. Both the formal and the informal "gatekeepers" in a system should be involved. It appears as if Sanders did not do much to discuss his proposed research with the key nurses prior to his contacts with the IRB. The nurses may have some valid reasons for suggesting that consent should be elicited in a different manner. Why not listen to the concerns and questions they have? Their fears may go beyond the perceived risks to the informants.
2. It is difficult to understand why this researcher should want to go back to the IRB to "attempt to educate them about social research." This is not his role in the organization. What is important, at this early stage in his research, is to see how the consent form was designed. It is said that the document posed risk to the women because it would document that "addicted women were engaging in illegal behavior." As I understand this case, the design involved comparison between the two groups. The form should have been worded so that no one would know from the form whether the woman in the research was addicted or not. The form should not be a document to identify any of the women as addicted or as engaging in illegal behavior. This is a standard procedure in this type of social research. Is Dr. Sanders aware of this?
A related question that no one has raised is whether the form has beenin a way that the participants can understand its contents. Assuming that the participants in the study can read and write, was this issue considered?
3. We are asked if the researcher should "protect research subjects" and not inform the IRB about the concerns raised by the nurses. My discussion in (1) and (2) addresses these issues.
4. I do not see the problem as one where the researcher is asked to "condone deception." I have suggested, rather, that we examine the process through which the researcher has established the principles that guide his research. This is a process that, in summary, involves not only a focus on the formal aspects but also the informal aspects of social relations in the settings where research takes place.
Prior to IRB approval, a researcher needs to understand the system within which he works and he needs to invite the cooperation of the grass roots personnel whose time will be taken up doing him favors associated with his work. I have discussed this process in an article written jointly with Leila Calhoun Deasy in "Patients in Programs at Area C Community Mental Health Center" (Washington D.C.: Department of Human Resources, 1971). With the collaboration of the nurses and possible advice of community persons from the area where informants reside, Dr. Sanders should be able to revise the form, if necessary. Since he is only two months into the study, it appears feasible to go back to persons already contacted and eventually obtain uniform consent from all participants.
5. A final point has to do with the settings for this research. It is difficult to tell whether the hospital and public health clinics are part of the same system or not. If they are not, Dr. Sanders needs to address the question of permission and consent of informants in both types of settings.
Table of Contents / Back to Chapter 4 / On to Chapter 5
About AAA / Join AAA / Jobs & Careers / AAA Meetings / AAA Publications
Sections & Interest Groups / Staff Directory / Anthro Links / Support AAA
or comments? We want to hear from you!
© 1996-2006, American Anthropological Association
2200 Wilson Blvd, Suite 600, Arlington, VA 22201; phone 703/528-1902; fax 703/528-3546