Responding to emergencies and sharing data
Apparently, Medecins Sans Frontieres (MSF) is what happens. MSF -- a Nobel prizewinning humanitarian organization -- collects data on the people they are assisting in the many emergencies they are engaged in. Not that this is a bad thing: some of the data collection (on routine medical procedures, adherence to treatments, etc.) are primarily for quality control purposes, and MSF also conducts operational research among groups affected by neglected diseases in impoverished countries. But data collecting in times of crisis leads to an ethical challenge: on the one hand, if the data collected could be used in beneficial ways by other researchers or agencies, it ought to be shared. On the other hand, the kinds of data collected in extreme circumstances (think sexual violence in the Democratic Republic of Congo or injection drug use in Thailand) is sensitive and could harm already vulnerable populations if it is not handled appropriately.
A recent article in PLoS Medicine by the former MSF International President Unni Karunakara describes how the organization developed a policy to resolve these ethical tensions. Much of it seems sensible and thoughtful, but one point might raise eyebrows. By sensitive data, MSF understands " ... any subset of data that could be misused against the interests of individuals whose data is included in the dataset or against MSF, or put either individuals or MSF at risk for political, financial, or other reasons." Apparently if MSF determines that the risk to its own organization is too high, then it will not share a dataset with others. But what if access to that 'MSF-compromising' dataset might benefit communities? I have pointed out elsewhere that MSF zealously protects its own reputation, sometimes at the expense of faithful disclosure of its own practices. Keeping a positive public relations profile is important for an organization largely dependent on donor contributions. Hopefully the tendency towards self-protection will not have an adverse impact on its data sharing practices.