Monday, February 12, 2007

The ethics of disaster research

At first glance, it does not look good: a disaster occurs in a resource-poor country. The response, usually bankrolled by the richer nations, partly consists of humanitarian aid and partly ... research involving humans. In addition to receiving the standard blankets and food packages, disaster victims may be enrolled in research, which can range from clinical trials involving medical interventions to social science research involving focus group discussions. Is research involving disaster victims inherently unethical? Is such research all a piece with the voyeurism of disaster media coverage, where suffering or dead bodies of the global South are displayed for visual consumption by the global North?

Stakeholders in the developing world seem to take a more sophisticated view than the knee-jerk accusation of exploitation. For one thing -- as an article by Sri Lankan researchers in the Lancet pointed out a couple of years ago -- some disaster research has revealed that some well-intentioned psychological counseling offered to disaster victims disasters is useless or even harmful. Disaster research is like emergency research: the predicament of the participants may open the door for exploitation, but such research is also useful and could be ethically conducted if appropriate protections are proposed and enforced.

Sri Lanka seems in fact to be taking the lead in the ethics of disaster research. The Institute for Research and Development last month organized an international conference with the title, 'Disaster Related Research Ethics: Developing world Perspective'. For a welcome change, 'international' here meant participants from south-east Asian countries affected by the recent tsunami were invited, rather than the usual suspects from the Eastern United States and Geneva. Members of the conference have formed a working group that aims to craft guidelines of the ethics of disaster research. Standard ethics guideline manufacturers -- such as the WHO or UNESCO -- should take note.