Sharing individual patient and parasite-level data through the WorldWide Antimalarial Resistance Network platform: A qualitative case study.

BACKGROUND

Increasingly, biomedical researchers are encouraged or required by research funders and journals to share their data, but there's very little guidance on how to do that equitably and usefully, especially in resource-constrained settings. We performed an in-depth case study of one data sharing pioneer: the WorldWide Antimalarial Resistance Network (WWARN).

METHODS

The case study included a records review, a quantitative analysis of WAARN-related publications, in-depth interviews with 47 people familiar with WWARN, and a witness seminar involving a sub-set of 11 interviewees.

RESULTS

WWARN originally aimed to collate clinical, in vitro, pharmacological and molecular data into linked, open-access databases intended to serve as a public resource to guide antimalarial drug treatment policies. Our study describes how WWARN navigated challenging institutional and academic incentive structures, alongside funders' reluctance to invest in capacity building in malaria-endemic countries, which impeded data sharing. The network increased data contributions by focusing on providing free, online tools to improve the quality and efficiency of data collection, and by inviting collaborative authorship on papers addressing policy-relevant questions that could only be answered through pooled analyses. By July 1, 2016, the database included standardised data from 103 molecular studies and 186 clinical trials, representing 135,000 individual patients. Developing the database took longer and cost more than anticipated, and efforts to increase equity for data contributors are on-going. However, analyses of the pooled data have generated new methods and influenced malaria treatment recommendations globally. Despite not achieving the initial goal of real-time surveillance, WWARN has developed strong data governance and curation tools, which are now being adapted relatively quickly for other diseases.

CONCLUSIONS

To be useful, data sharing requires investment in long-term infrastructure. To be feasible, it requires new incentive structures that favour the generation of reusable knowledge.