Big data is a big concept to wrap your arms around. But once you gain an understanding of what you can do with your data, the results can be transformative.
At the 2013 HIMSS Annual Convention and Exhibition on March 4, a health system CIO and a clinical strategic consultant defined big data as a collection of data sets so large and complex that it become difficult to process using standard database management tools. In healthcare, it comes into play when you consider all the pieces of information about every patient your organization encounters over their lifetimes.
Gregory Veltri, CIO of Denver Health, and Mical DeBrow, PhD, RN, co-presented "Extracting Value from Healthcare Big Data with Predictive Analytics," an advanced-level educational session.
The presenters pointed to an ATLAS study on the value of tamoxifen therapy as an example of an analytics "win." The study showed that a 10-year regimen of tamoxifen – versus the current recommendation of five years – produced a 56 percent reduction in recurrence of breast cancer.
However, DeBrow warned, "Technology alone cannot resolve the management of data. Technology will provide the data," he added, but healthcare organizations will have to make a "cultural shift in how data is perceived and managed…It is important to remember that this is an integrated effort that requires clinicians of all types, as well as non-clinicians, to get this right."
Veltri provided the following steps toward achieving effective data management:
- Define measures in a data dictionary.
- Collect the desired data.
- Measure and analyze the data against known standards and benchmarks.
- Make the data actionable.
It all boils down to this, according to Veltri and Debrow: Turn your data into knowledge, and your knowledge into action.
"If we don't drive our healthcare businesses off of actionable data," commented DeBrow, we're not going to be here much longer."
The ultimate goal of healthcare data must be to predict the behavior of the system and its consumers, the presenters said.
Veltri characterized big data management as a process of creating "one source of truth." He added, "What that means is creating one definition for every data element that you have – and it must be understood by your users. A data dictionary that is relevant to the users is an essential part of what you do."
Photo attributed to Procsilas Moscas via Creative Commons license.