EULAR calls for privacy, transparency when using big data in RMD research
Privacy and transparency are among the core points to consider when utilizing big data in rheumatic and musculoskeletal disorders (RMDs), according to the first EULAR-endorsed recommendations for use of big datasets in research.
The use and implementation of big data provide opportunities for better patient care and clinical decision-making, yet no academic societies have thus far developed guidelines to guide practice and ethics until the EULAR recommendations.
EULAR’s “points to consider” were drafted by a multidisciplinary task force of 14 international experts to synthesize essential issues surrounding big data acquisition, obtain a consensus, and provide a framework that will guide the collection, analysis, and use of such in RMDs. [Ann Rheum Dis 2019;doi:10.1136/annrheumdis-2019-215694]
The term big data may be widely used but there is not one commonly accepted definition, said the taskforce; hence the first overarching principle defines big data. “The term ‘big data’ itself has a broad definition and could define the large-scale datasets that include imaging data, electronic health records, or administrative claim records, among others,” said principal author Dr Laure Gossec of the Sorbonne University in Paris, France. “Big data is also used to refer to specific analytics and statistical methods, such as artificial intelligence and machine learning.”
Ethical issues related to privacy, confidentiality, and transparency when using big data should be considered, said the taskforce, even as it highlighted the unprecedented opportunities big data could provide in terms of transformative discoveries to improve RMD research and practice. The ultimate goal of utilizing big data in RMDs is to improve the health, lives, and care of patients, including health promotion, assessment, prevention, diagnosis, treatment and monitoring of disease, the taskforce added.
Some points to consider are the need for data harmonization, including the pooling of existing and future datasets to facilitate data interoperability; big data should be based on the FAIR (findable, accessible, interoperable, and reusable) principle; open data platforms should be preferred; privacy by design should be applied to the collection, processing, storage, analysis, and interpretation of big data which should be underpinned by interdisciplinary collaboration of scientists, relevant clinicians/healthcare professionals, and patients.
Other salient points include explicit reporting of the methods used to analyse big data in scientific publications; benchmarking of computational methods for big data used in RMD research; independent validation of conclusions and/or models drawn from big data prior to implementation; proactive implementation of findings in practice; and interdisciplinary training of clinicians and scientists on big data methods in RMDs.“The use of big data … is a rapidly evolving field with the potential to profoundly modify RMD research and patient care,” said Gossec. “These first EULAR-endorsed points to consider address core issues including ethics, data sources and storage, data analyses, artificial intelligence, the need for benchmarking, adequate reporting of methods, interpretation, and implementation into clinical practice. We hope these points to consider will promote advances and homogeneity in the field of big data in RMDs and provide a useful guidance to other medical disciplines.”