Re-post from
NEW ORLEANS—When considering your patient data, “it doesn’t help to know what happened 30 days ago. It’s too late,” said Mical DeBrow, PhD, RN, Siemens Clinical Strategic Consulting, speaking during a March 5 education session at the Health Information and Management Systems Society (HIMSS) annual convention.
DeBrow encouraged the audience to “push on your vendors to expose data models. It’s going to be important in this entire industry.” With the expected growth of patient data, datasets will be so large that it becomes difficult to process them using standard database management tools. “Every piece of data and every result for the entire lifetime of the patient times 315 million people in the U.S.” is resulting in a storage problem since we are beyond pentabytes. “We’re collecting every piece of data but we’re not doing anything with it.”

But it is important to get comfortable with predictive analytics because “analytics wins,” he said. For example, a study comparing 10 years of tamoxifen to the current recommendation of five years on the drug for breast cancer patients resulted in a 56 percent reduction in the recurrence of breast cancer. “That should change practice.”

The data avalanche is here, said DeBrow. “The question is, are you buried in it? There are mountains of data you cannot get to. Technology alone cannot resolve management of data. A cultural shift in how data are perceived and managed is required.” Data management will change the way healthcare organizations work. “If it doesn’t, we’re going to shoot ourselves in both feet.” But, it’s important to remember that this is an integrated effort, he said, that requires clinicians as well as nonclinicians to get it right.  

“IT by itself will fail at this endeavor,” said Gregory Veltri, CIO of Denver Health. “We aren’t clinicians. You need people distanced from IT that drive clinical practice” to be involved. Providing data to patients that predicts their future outcomes could make a difference, he said. “We need to bring data home to the patient and make it part of their lifestyle. You’ve got to make the argument that’s beyond ‘because it’s the right thing to do.’”

Analytics is resource intensive, costly and forever, Veltri said. “Analytics is going to be a forever expense. It takes very talented people who are hard to find and it takes clinical commitment.”

A data dictionary is critical to successful data management. “You need one layman’s definition for every data element,” said DeBrow. That’s essential because “you’re not comparing like to like unless you have those definitions.”

Make the data actionable, DeBrow said. “If we don’t drive our healthcare businesses off of actionable data, we’re not going to be here. Turn data into knowledge and knowledge into action.” To know data you have, you must assess against standards and benchmarks for retrospective analysis. “Looking prospectively requires predictive analysis. Do a data map,” he said.

Meaningful and actionable insights come from predictive analysis, Veltri said. “Predictive analytics drives competitive advantage.” For example, Denver Health found that shifting from the highest costing blood thinner to the lowest makes no difference in clinical outcomes but saved $180,000 a week per floor. “That was just based on data. You can make that difference in your organization,” he said.

Having one source of truth is important, said Veltri. “We dumped all our data in and didn’t define it. That will come back to haunt you as it has us.” One definition for every data element you have has got to be understood by users, he said. “We now have millions of elements that are undefined and it takes programmer after programmer to deal with that.”DNA analysis and genomic surveillance are being used to predict best treatment plans, specific drug value and interventions prior to the development of chronic illness, DeBrow said. “Are you ready for data as a disruptive technology? It distracts you from everything else you’ve got to do but if you don’t do it, what are your options?”



Comments are closed.