The healthcare world is swimming in data, from insurance claims to electronic health records to patient monitoring devices to the bar codes on every IV bag and vial. Analytics, the software containing the algorithms that try to make sense of data and improve patient care, are becoming a standard part of the healthcare IT toolkit. But do healthcare systems still need actual human analysts in addition to this technology?
It depends, says long-time healthcare CIO Terry Carroll, who works with the SPRING Network, an organization devoted to organizational change. “Not every organization can have this kind of capability," he says. Smaller or rural hospitals, for example, need to devote their limited resources to day-to-day delivery of care. But healthcare systems with more money and more diverse missions—and especially those that do research— may need human beings to make sense of data and explore ways to apply key findings.
Joe Dudas, division chair of Enterprise Analytics at the Mayo Clinic, thinks of analytics deployment in three phases: discover, translate, and apply. Analysts work in all three phases, but particularly in the “translate" phase, taking the insights produced by analyzing the data and working with users to move them into action.
Your organization may use an analyst, or a group of analysts, to help with:
It's vital for lab managers to find a way to participate in enterprise-wide analytics projects because the data they can contribute may not be readily understood by analysts who don't have specific expertise in lab matters, explains Peter Gershkovich, director of pathology informatics at Yale University Medical School. "The role of labs should increase," he says. "Analysts don't understand the complexity of the lab data model, and that prevents them from adequately addressing data requests, either for research or for quality improvement."
Organizations trying to establish an analyst function should think carefully about how to position it in the community, notes Ed Hammond, director of the Center for Health Informatics at Duke University. The best analytics in the world won't change anything unless they're connected to the people who need the insight.
In a study of one department at Duke, clinicians simply weren't using the information provided by the analyst group—except for one researcher who needed material to make slides for a talk. “When you're doing analytics, who are you doing it for?" Hammond asks. “Is it only the research community or is it the clinical community, as well? How do you make the analyst part of the team and get the information to the right people? How much is the analyst responsible to educate the potential users? It's a lot of questions, but you have to ask them to move the analyst function into a broader, more useful spectrum."
Next-gen analytics: Here's what's coming in the future
By Bill Siwicki
Hospitals should expect orders of magnitude more data – but will also see emerging tools such as artificial intelligence and 5G connectivity helping to put both structured and unstructured information to work.