Data sharing key to creating a successful learning health system

Sharing data to create a learning health system will require both a change in culture and new technology. However, despite the high potential for such a system, incentives often are not in line with requirements to make this happen, says Yale School of Medicine cardiologist Harlan Krumholz.

According to Krumholz, Sharon Terry of the Genetic Alliance and Joanne Waldstreicher of Johnson & Johnson--writing in a viewpoint article published online this week in the Journal of the American Medical Association--those in possession of data, including health systems, researchers and others, generally bear the costs associated with data sharing. It’s difficult, they say, to fully integrate data from various sources, to fully analyze and to apply it.

Among the strategically important aspects, then, will be ensuring patient access to their data, which may require new tools and infrastructure as well as regulatory changes,the authors write.

Three priority considerations, they say, are:

  1. Fostering a culture of data sharing: Krumholz, Terry and Waldstreicher advocate financial incentives to reward data sharing and interoperability, with penalties for not doing so. They also promote acceptance of the need to reuse participant-level data.
  2. Creating additional operational functionality: The authors urge government to establish the legal and regulatory pressure for data sharing, including requiring it for studies that use public funds. They advocate continued efforts to standardize application programming interfaces (APIs) and technology that helps patients control movement and access to their data.
  3. Building capacity for continuous improvement: Since many of these suggestions already are under government purview, the authors suggest a White House Initiative to bring all the efforts together and use of benchmarks to determine progress and effectiveness.