Path forward for NIH big data initiative will be challenging, researchers say

How to create better value from the data contained in electronic health records is just one of the challenges the National Institutes of Health is diving into with its Big Data to Knowledge (BD2K) initiative.

NIH announced the initiative in May to focus on gaining better knowledge from existing data.

Its path forward involves engaging a range of stakeholders--including biomedical researchers, clinicians, citizens, data scientists, publishers, libraries and others--on their unaddressed needs and the necessary policies and practices involved in the use of biomedical big data, according to an article published at the Journal of the American Medical Informatics Association.

While developing technology to address specific high-need challenges for researchers is a first step, technology alone is not the answer, the authors state.

The initiative also focuses on training big data practitioners with skills, including computer science, mathematics and statistics, biomedical informatics, biology and medicine, among others. Interdisciplinary teams to design studies and perform the subsequent data analyses will be required to address the complex questions being posed.

A sustainable and scalable plan also calls for data-discovery capabilities allowing researchers to find and effectively use relevant biomedical data. In the initial efforts in creating a Data Discovery Index (DDI), stakeholders will be encouraged to learn from other fields and conduct short-term pilots on the best way to index existing datasets.

EHR data poses special challenges because it's captured outside of research settings and must be repurposed to be used in research, according to the authors. It poses special challenges with consent and protection of patient privacy that may require changes in policies and practices. The initiative also is focused on how to best go about making these changes.

Princeton University experts argued recently that privacy has the potential to crash big data before there's a chance to get it right, and finding the right balance is key to future success.

To learn more:
- read the JAMIA article