03 August 2008

Ethics and knowledge practices in HIV/AIDS reserach in the medical and social sciences

It is difficult to know where to begin a "blog."

So let this be a beginning. There are a number of things that I will address in this blog. Most have to do with various aspects of politics, culture, health and sexuality in southern Africa.

Since my book, Unimagined Communities is due soon from the University of California Press, and since I am now the new a reluctant chair of the Witwatersrand University Human Research Ethics Committee, the issue of ethics and HIV/AIDS research is one of the issues that I will begin this blog with.

Differences between the social and the medical sciences:

Fundamental differences exist between the social sciences and the bio-medical sciences in the assessment of ethical standards for research on sexuality and HIV/AIDS, and in the conduct of research on these topics.

These differences have serious implications for the application of ethical standards for University ethics committees in their assessment and clearing of research proposals, for the nature and quality of the knowledge that this research produces.

This also has significant impacts on the formulation of public health policy, for the effectiveness of public health programmes and interventions aimed at improving reproductive health and the prevention of STIs including HIV.

These differences arise primarily from

  • epistemological differences, that is, differing standards for the assessment of ‘truth’ or validity of knowledge,
  • methodological differences, that is, different ways in which data is gathered, and,
  • differences in what we might call, more generally, knowledge practices.

The epistemological and methodological difference are already well known and extensively discussed in literatures for all fields concerned. The latter, however—knowledge practices—are a set of practices (activities, procedures, customary usages deriving from intellectual traditions of disciplines) aimed at gathering and interpreting knowledge, rather than (or in addition to) the usually conceived methodological and epistemological concerns of standard science and the philosophy of science.

The notion of ‘knowledge practices’ includes issues of

  • distribution and creation of knowledge,
  • knowledge management,
  • knowledge transfer and communities of knowledge,
  • ‘traditional knowledge’ and ‘indigenous knowledge systems’.

The idea of knowledge practices has emerged as a central focus in anthropology, especially in areas of material culture, anthropology of science and technology, and in the new histories of science deriving from work by Bruno Latour, Michel Foucault, and others. It is concerned with the social embeddedness of ways in which knowledge is created, used, distributed and evaluated, especially by specific communities of knowledge practitioners (educators, scientists, technicians, bureaucrats, doctors, healers and others).

One example of a ‘knowledge practice’ in the areas of reproductive and sexual health is the ‘global gag rule’ imposed by the US government on organizations involved in reproductive and sexual rights and health issues.[1] The US government, through it agencies such as USAID and the State Department require that any organization that received funding from the US government may not discuss in any way issues of prostitution, transactional sex and abortion, and must promote ‘abstinence’ only programmes.

This is a political imposition of a specific moral or moralistic approach to intervention in matters sexual and reproductive, and severely limits effective activity in these fields by restricting the kinds of knowledge that can be communicated to the public. It also restricts the kinds of knowledge that may be deployed in development of search for new knowledge (research agendas, questions, methodologies) and has had extremely deleterious effects on the ability of NGOs that might be funded—directly or indirectly— by the US government to be effective in their overall mission.

For instance, in directly prevents distribution of condoms, knowledge about reproductive and sexual alternatives, and any interaction with abortion or prostitution. Since the US, through USAID and the State Department, is the largest global funder of development programmes worldwide, its impact has been enormous. This is not an issue of epistemology or methodology, but one that can be more generally considered under the rubric of knowledge practices.

A critique of this issue arises from the interaction of bio-medical research practices and ethnographic research practices in the areas of sexuality and reproductive health. Because the biological, medical and social-cultural aspects of sexuality and reproduction are virtually inseparable, bio-medical and social scientific approaches to research in these areas come into close contact. Ethical and moral questions are necessarily paramount in all matters of sexuality and reproduction. Conflicts between the two scientific paradigms (bio-medical and social scientific), therefore, come into sharpest focus here.

Institutions and research programmes, therefore, spend a great deal of effort on resolving ethical issues having to do with these research projects, but in our view these often result in misunderstanding and talking past one another.

Especially in anthropology, and in the case of ethnographic methods used in many of the social sciences today, knowledge is derived from its distinctive method of ‘participant observation’, or observant participation in communities. When we conduct research in sexuality and reproductive health, anthropologists—as a community of knowledge practitioners—are involved in specific knowledge practices (ethnographic methods) that are designed to interact with the knowledge practices involved in sexuality and reproduction as close to the ‘ground’ of actual practice as possible.

The randomised control trial (RCT) versus the ethnographic approach

This often brings it into conflict with the bio-medical approaches to knowledge creation (‘research’) that also aim to produce knowledge about sexual and reproductive practices, but do so in an entirely different way. Overwhelmingly, research in the bio-medical sciences is based on one form or another of the randomized controlled trial (RCT). The RCT is a highly focused, rigorous, rule-bound and ‘objective’ empirical method involving

  1. randomized selection of
  2. research subjects from a
  3. population for whom different
  4. experimental interventions or
  5. invasive procedures are allocated to
  6. two or more groups (control and experimental group).

This method is held to eliminate both observer and subject bias (the so-called ‘double-blind’ research design), ‘spurious causes’ and ‘confounding factors’.

For medicine, and especially for pharmaceutical companies, this is especially important with respect to legal challenges, profitability, and political acceptability. Since health, especially public health, is understood to be concerned with the biological state (health, illness, disease, epidemic infection) of population composed of biologically-defined collections of individual subjects, who are themselves either susceptible, immune, recovered (or dead), this does not generally concern itself with the social causes or context of illness. These are in fact specifically excluded as ‘confounding factors’, or bias.

All experimental interventions, including questions to be asked and expected responses are known in advance. Unexpected results, including novel responses or ‘creative interpretations’ of experimental conditions, questionnaires, interventions and informed consent protocols, are regarded as experimental errors, ‘confounding factors’, ‘data outliers’, experimental bias, and are usually dropped from consideration. From the point of view of social sciences, this data is lost, while from the perspective of RCT methodology, it is regarded as excess, epiphenomenal, or external.

In the social sciences, however, the so-called confounding factors and ‘bias’ are understood as the more fundamental issue, while the biological ‘facts’ of illness and disease are secondary to their social and cultural contexts. Unusual responses, creative or alternative interpretations.

Above all, for bio-medical sciences, all such experimental (RCT) or studies are regarded as potentially or actually harmful or deleterious to the subjects of the research. In order to make this type of research morally acceptable—and thus, ‘ethical’—it is held that subjects must

(a) be fully informed of the potential risks (and possible benefits—though these do not usually accrue to the experimental subject, but rather to the population) of the RCT,

(b) give specific and informed consent in a formal and legally-valid way, usually by signing a written form after they have listened to a statement about the nature of the research and their participation in it, and

(c) receive, in some cases, some form of compensation in goods, money, or services (such as counselling or other medical services) to offset the real or implied risk to their own health and/or well-being or, minimally, for their time (as transactable commodity).

The research subjects’ acceptance of gifts received as ‘compensation’ also legally guarantees their consent in the form of an implied exchange contract in which consent is freely ‘sold’ on the basis of full knowledge of its value in exchange for other goods and services provided by the experimenter or institution under whose auspices the RCT is carried out. These practices are now fully routinised, tested in international and national legal forums, and validated by most global scientific organisations.

Implicit in this methodological division is a more general issue of the knowledge practices of the research situation. For social scientists engaged in ethnographic methods, the object of their interest is the knowledge and beliefs concerning health and illness.

In the areas of sexuality, this involves knowledge and experience of sexual practices, feelings, identities, and emotions, and the way they are embedded in social and cultural structures, especially those involving gender and social power.

For reproduction, this involves myriad issues around kinship (affinity, consanguinity, ‘blood’, solidarity, family, wealth…), masculinity and femininity, fertility and many other ‘embodied’ concepts of self, other, group, identity and so on. Issues of health, well-being, illness, disease, and death are fundamental to all of these concerns.

For the social scientist, these are held to be integral to being human, while for the bio-medical sciences and RCT, humanity is—experimentally speaking—a confounding factor, and legally speaking, something that must be managed through informed consent protocols.

Ethnographic methods versus the Randomised Control Trial

Ethnographic methods in particular and social science methods more generally, do not operate with the same methodological assumptions implicit in RCT.

First of all, the research situation is not held to be present risks to those we work with that are not also present for the researchers themselves. In other words, the ‘risks’ that exist are risks that are common to all of us as human beings, rather than specific to the subjects of research under investigation by an ‘objective’ scientists. Social science ethics, and anthropological ethics in particular, assume a community of risk that includes the researcher. Thus, the anthropologist is ethically enjoined to prevent harm to those he or she works with, but also may be expected to mitigate external risks to the communities involved.

Paradoxically, these external risks might come from RCT experiments, especially those run by large pharmaceutical industries, or by First world bio-medical researchers in Third World situations. Thus, the social scientist may come into direct ethical conflict with bio-medical research, especially with respect to RCT.

Recent examples include studies RCTs involving circumcision in southern Africa and microbicide trails. Microbicides once looked like they would offer protection from HIV transmission. Although they should have worked, given everything we know, they did not. It seems, quite simply, that we did not know enough. We do not precisely how HIV enters the body; assumptions were made that proved unwarranted. There was little investigation of the erotic effect of these gels given the specific beliefs and practices of those who might use them. Regardless, because they looked effective in vitro, it appeared that they would be effective in vivo, 'in life', as we say. But 'in life' was not a meaningful category: rather, sexual lives are lived through culture, and it was the cultural categories that interfere. In some RCTs, the 'subjects' were subjected to testing that actually increased their risk of infection. though the trials were stopped, little was done to investigate the cultural context in which 'in vivo' actually has meaning and real effect.

With respect to institutionally mandated ethical protocols, however, anthropologists do not generally believe that their knowledge practices present unusual risks to those they study. This presents a special challenge to expectations that formal informed consent protocols are required in all cases of research.

[1] According to the Council on Anthropology and Reproduction, Society for Medical Anthropology (a unit of the American Anthropological Association), ‘The Global Gag Rule prohibits foreign NGOs that receive U.S. family planning assistance from offering abortion, counseling about abortion services, or lobbying to change abortion laws, even if they do so with non-U.S. funds. The policy was first enacted by President Ronald Reagan in 1981, rescinded by President Bill Clinton in 1993, and reinstated by President George W. Bush in 2001. The Global Gag Rule puts the United States at odds with global agreements on reproductive rights, including those reached at the International Conference on Population and Development (Cairo, 1994), and the ICPD+5 conference (New York, 1999. Online: http://www.medanthro.net/stand/overview/index.html (accessed 21 Feb 2008). See also http://www.globalgagrule.org/background.htm