This paper was prepared to inform the Chatham House workshop on ‘Data Sharing in Disease Surveillance: Experiences and Vision’, held on 12–13 February 2014, and was based on interviews with more than 25 stakeholders in disease surveillance data sharing, representing primary data producers, secondary users, research funders, international organizations and other relevant groups.
Many data sharers agree that the quest to normalize data sharing in disease surveillance needs to proceed carefully and relatively slowly in order to successfully navigate the sensitivities and resistance currently being shown, and that the first thing is to recognize people’s fundamental interests and address them.
Many recommend proceeding with a stepwise and mixed approach; several data sharers say that pressing immediately for open access to all data in disease surveillance is likely to result in failure.
Some say the first step should be to identify the impediments to widespread data sharing. One suggestion is to convene an international group – with representation of the different data fields and barrier groups in both developing and industrialized countries, to engage in a process to articulate the case for data sharing and produce guidelines or a framework for doing so in disease surveillance.
Other suggestions include:
- Developing methods for evaluating data-sharing initiatives, to leverage the aspects that are characteristic of success.
- Articulating a fairness statement on data sharing between countries and between researchers.
- Drafting by funders of a model agreement, or set of agreements, that primary data producers and secondary users can use to spell out the roles and responsibilities on both sides. This could allay some of the fears among data producers that funders want to wrest the data out of their hands to give to others to work on. It would enable funders to show they recognize the larger role that producers want to play and could help producers determine what would be legitimate to ask for in their agreements with secondary users.
- Examining successful initiatives to illustrate the benefits and impact of data sharing, derive best practices and determine the value of conducting more aggressive pilot studies of data-sharing initiatives.
- To make data more shareable, not embarking on, or funding, data collection that is not going to result in analysis, interpretation and reporting.
As for what data sharing in disease surveillance might look like in the future, many believe it is likely to remain fragmented, owing to the diversity of the data types involved and varying skills of those who collect it. Some natural groupings might emerge to simplify the picture, but it is considered that those are likely to be disease-specific.
Another factor that could have an impact on the current model of data sharing is the advent of mobile phone health apps that allow citizens to send data to researchers and participate more actively in surveillance. It was argued that app users have shown extraordinary willingness to share their health data, raising the question of whether such interactive surveillance could become an increasingly important source of health surveillance data in the future.
Whatever the future model of data sharing, some experts say that secondary users may always feel that they are not getting enough data, quickly enough, in good enough shape. However, it was argued that data-sharing practice is moving in the right direction. It may not be progressing as fast as desired but there is greater openness and insight into the need for sharing than there was 10 years ago and it may be reaching a tipping point.
But for many, the ultimate vision for data sharing, and the goal it is trying to address – improved public health – is a world where epidemiologists, public health workers and investigators can ask and answer questions about health in their country without the work having to be carried out in external organizations. The best policy, some say, is to strengthen decision-making systems and public health career structures in developing countries so that talented people in low-income countries want to make their careers in analysis of health information and countries have the capacity to plan for themselves, decide their own priorities and address their own problems.