machine learning, health care

At a time when the amount of data being generated continues to soar and the data itself is increasingly distributed, the ability of organizations to share data within their own operations and with other companies is critically important. In the modern business world, data is crucial to operations and decision making, and the faster and more easily the data can be shared and consumed, the faster those informed decisions can be made.

That’s particularly true in the health care industry, where the ability to share data – including from such systems as electronic health records (EHRs) – can help inform decisions on everything from patient treatment to insurance information to payment. The challenge is ensuring that the data from one health care system can be integrated into another.

It’s what is driving the push in the health care industry to create an environment where data can be used in a coordinated way, according to Symplr. For health care executives, that means enabling greater accessibility to better health care and reducing costs for both patients and providers.

It’s “the ability of different information systems, devices, and applications to access, exchange, integrate and cooperatively use health care data in a coordinated way,” the health care operations software maker wrote in a blog post last year. “It’s the reason why patients can send health care data from their Bluetooth-enabled blood pressure cuff to their provider’s electronic health record (EHR). It’s how an emergency room (ER) physician can access test results from a patient’s cardiologist with the click of a button. Or why a primary care physician can directly send electronic care summaries to a gastroenterologist when referring patients. It’s also a driving force behind a payer’s or health system’s risk management and compliance.”

More Data Insights Needed

Different organizations may have access to the same data, but unless the right systems, content and privacy standards and processes are in place in a coordinated manner, they won’t be able to pull important insights from it.

The job of the Office of the National Coordinator (ONC) for Health IT, which is housed in the U.S. Department of Health and Human Services, is to create standards for patient information, and it wants to reach nationwide health care interoperability by next year.

Databricks, which provides a data lakehouse architecture for storing and accessing data, outlined the problem in a blog post in May, saying that “high quality and usable data are the lifeblood of any advanced analytics or machine learning system, and in the highly complex health care industry, data quality (DQ) has historically been poor, with limited standards and inconsistent implementations.”

A broad range of health care organizations, IT vendors and agencies are working to address the issue of data usability in health care.

Enter The Sequoia Project

One of the groups pushing for such data usability is The Sequoia Project, a non-profit heath IT interoperability organization, founded in 2012 that now has more than 80 member organizations. The group last year created its Interoperability Matters Data Usability Workgroup, which in December published its data usability guidance report, which outlines improvements needed for what it calls “semantic interoperability of health information,” starting with the quality of the clinical data shared between health care providers.

The guidance “can be readily adopted by EHR and health information exchange vendors, implementers, networks, governance frameworks (i.e., ONC Trusted Exchange Framework and Common Agreement (TEFCA), Carequality, eHealth Exchange, CommonWell, etc.), and testing programs,” the authors of the 41-page document wrote. “This guidance includes data systems and processes from the originating EHR through intermediaries to the end user.”

In July, the organization – in collaboration with the American Health Information Management Association (AHIMA) – launched an initiative called Data Usability Taking Root to make data more usable in the industry. Others participating in the launch included Epic, Civitas Networks for Health, HIMSS Electronic Health Record Association (EHRA), and New York eHealth Collaborative.

The goal is to bring together organizations to implement the steps outlined in the usability guidance report

“Over three years, more than 260 health organizations worked together through The Sequoia Project to develop practical guidance to make health data more useful for health care providers, health IT vendors, public health, health information exchanges, and patients,” Sequoia Project CEO Mariann Yeager said at the time. “It’s time to put this guidance into action for the public good.”

Reaching the Summit

Since then, organizations in the Taking Root effort worked on areas important to them, such as data provenance and traceability of change, data integrity and trust, or data tagging and searchability, according to Didi Davis, the project’s vice president of informatics, conformance and interoperability.

The work included a series of virtual events over the summer. Earlier this month, many of the participants gathered in Washington DC for the Data Usability Taking Root Summit.

According to Sequoia Project officials, an array of organizations attended the summit and dozens are participating in activities to make the data being exchanged in the health care industry better and more accurate.

“Data usability at its root is about the availability, completeness, clarity, and portability of information,” said Duncan Weatherston, CEO of Smile Digital Health, a summit co-sponsor. “We have to start with data – and the usability of data – to achieve goals like accessibility for people everywhere, innovations that extend longevity and quality of life, and bending the cost curve of care.”