Clinicians Need Unconventional IT Tools For Toughest Cases

5 Key Elements For Clinical Decision Support Systems

Once upon a time, the term disruptive technology caught people's attention because it suggested original thinking and innovation. The expression has lost its luster, but there are still innovators out there who know how to think this way.

I've written recently about the importance of basing a clinical decision support system on strong medical research derived from randomized controlled trials whenever possible. But suppose there are no RCTs available to meet clinicians' needs? What kind of guidance do you provide your doctors in such situations?

In the past, physicians have relied on expert opinions and other, less-rigorous data to make their decisions in such situations. But with the emergence of natural-language processing, advanced EHRs, and data warehouses, clinicians now have new options as they look for advice on diagnosis and treatment.

A recent case report from doctors at Stanford University School of Medicine illustrates just what these IT tools are capable of.

[ To find out which medical apps doctors and patients are turning to, see 9 Mobile Health Apps Worth A Closer Look. ]

Jennifer Frankovich, MD, and her colleagues admitted a critically ill teenager with systemic lupus erythematosus (SLE) complicated by kidney dysfunction and pancreatitis. When they searched the medical literature, they couldn't find any RCTs or other solid research to guide their actions.

So instead they looked at the data in the hospital's EHR and data warehouse, the Stanford Translational Research Integrated Database Environment. STRIDE's search engine let the doctors locate the records of 98 other pediatric patients treated for SLE. And among those also suffering from kidney complications and pancreatitis, the doctors discovered that the risk of blood clots was relatively high. That finding gave them reason enough to administer blood thinners.

Vendors such as MModal, Nuance, and Health Fidelity now offer the ability to sift through massive amounts of data by using natural-language processing, voice recognition, and other technologies to help clinicians make informed treatment decisions.

Dan Riskin, MD, a natural-language processing specialist and the CEO at Health Fidelity, has been trying to take data warehouses to the next level. During a recent phone conversation, he explained how the vendor's NLP platform, called Reveal, can help populate a warehouse with patient data that previously was unavailable--namely, the unstructured comments in the clinical notes section of e-records.

Reveal "allows a hospital to pass their unstructured clinical narratives through the NLP Web service and receive back clearly defined SNOMED and ICD 9 codes that map easily within their data warehouse," Riskin said. It allows for more robust analytics, improves accuracy, and reduces the need to do manual coding, he said.

Clinical analytics, data warehousing, NLP, data mapping: These aren't concepts the average physician learned about in medical school. But when a physician's years of training aren't enough to manage a difficult patient, it's time to take a fresh--disruptive--approach.

The 2012 InformationWeek Healthcare IT Priorities Survey finds that grabbing federal incentive dollars and meeting pay-for-performance mandates are the top issues facing IT execs. Find out more in the new, all-digital Time To Deliver issue of InformationWeek Healthcare. (Free registration required.)