Big data in healthcare needs governance
I recently had the pleasure of delivering the industry keynote for the California Health Information Association in Santa Clara, CA. Preparing was a good exercise for me to think about the practical aspects of governance, particularly in the context of skills and knowledge of the health information management (HIM) professional. As professionals who have managed information for decades, HIM professionals have a unique knowledge of documentation practices, privacy requirements, record retention requirements and quality management.
Here are some presentation highlights:
- The intense transformation pressure on all components of the healthcare ecosystem are driving collaboration, wellness and consumer engagement strategies like never before. A natural extension of collaboration is better use of all types of data through applying big data technologies.
- With 17.6 percent of the US GDP being consumed by healthcare spending, we must take bold steps to improve quality and reduce costs. These steps include increased data sharing, care collaboration (social and health) across boundaries and proactively identifying and mitigating risks in care delivery. See these examples of big data at work for more info: UCLA and the Hospital for Sick Children
- Population health activities are diverse and create data integration challenges. The activities range from identifying and placing high risk patients in care management programs to developing disease registries to study disease patterns and identify best practices. Data integration is particularly problematic because claims data, EMR data and historical warehouse data may be used; clearly this means inconsistent formats and standards (or lack thereof), and different purposes for the original data creation. Unless managed correctly, the end result could be the classic “garbage in, garbage out” which may jeopardize consumer engagement and trust.
Big data technologies truly create a world of the possible instead of the historically siloed data environment. However, with this new world, the governance challenges are exploding. And, addressing governance requirements takes a more agile, iterative approach. Simply put: “we don't know what we don't know.” Never before have we married structured data with the unstructured content that might include text documents, data streams from devices (such as glucometers and monitors) and social media data. Therefore, we are charting a new course that requires discovery, interpretation and adjustments.
The newly acquired unstructured data, whether used with the structured data or on its own merit, poses governance questions such as:
- Can the customer/patient data be used in an identifiable fashion?
- Should a limited data set or de-identification be applied?
- Should the data be masked during processing or at output due to privacy requirements?
- How long should the data be retained given state or federal retention requirements?
- How much trust can be associated with the information given the data sources and quality measures applied throughout the lifecycle of the data?
If you want to start your big data initiative on solid footing, engage a health information professional or hire a HIM consultant to guide you as you explore governance requirements in the world of the possible.
For further reading
- The American Health Information management Association (AHIMA), the national association for HIM professionals, recently conducted a study of information governance policies and processes now available for your reading and research.
- Analyst Report: Forrester Wave for Data Governance Tools, Q2'14