Sharing individual-level data from clinical and public health research is increasingly being seen as a core requirement for effective and efficient biomedical research. This article discusses the results of a systematic review and multisite qualitative study of key stakeholders’ perspectives on best practices in ethical data sharing in low- and middle-income settings. Our research suggests that for data sharing to be effective and sustainable, multiple social and ethical requirements need to be met. An effective model of data sharing will be one in which considered judgments will need to be made about how best to achieve scientific progress, minimize risks of harm, promote fairness and reciprocity, and build and sustain trust.

The abundance of South African clinical and public health research data has the potential to unlock important and valuable future advances in biomedical science. Amid increasing calls for more effective sharing of individual-level data, commitment to promote access to research data is evident within South Africa’s public research sector, but national guidance and regulation are absent. This qualitative study examined the perceptions, experiences and concerns of 32 research stakeholders about data-sharing practices. There was consensus about the utility of data sharing in publicly funded health research. However, disparate views emerged about the possible harms and benefits of sharing data and how these should be weighed. The relative dearth of policies governing data-sharing practices needs to be addressed and a framework of support developed that incentivizes data-sharing practices for researchers that are both ethical and effective.