How natural language processing can turn case notes into actionable intelligence

Finding ways to do more with health data was a major theme at this year’s EHI Live conference in Birmingham.

One of the best examples of this trend was a talk given by Paul Henderson of Ascribe on ways better data analysis can lead to better healthcare outcomes.

Healthcare providers often focus on the wrong aspects of Big Data solutions, he said. Raw computing processing power often draws the most attention, but clinicians should be looking for solutions that can help them identify meaningful patterns in historical data. “Old case notes are a seldom-used goldmine,” he said. By looking at trends over time, clinicians can turn their old case notes into actionable intelligence.

One of the major data challenges clinicians face is that some of the most useful healthcare data is difficult for machines to interpret. Doctors may use euphemisms or turns of phrase in their notes that can be difficult for computers to understand and put in the proper context. Providers need to use solutions that can be taught to process natural language, said Henderson. Over time, a system can be taught to recognize patterns in the grammar and diction of clinicians’ notes, making the data more accurate and more useful.

While teaching a machine to interpret natural language can be time-intensive, it can yield insights that can lead to improvements in care. He pointed to a recent Ascribe pilot program with Leeds NHS Teaching Hospitals where the hospital was able to identify geographic hotspots for health problems, such as binge drinking. This knowledge helps officials track how they’re providing services and allocate resources effectively. Rather than making assumptions about how best to serve their community, clinicians can change their organisation to suit the communities evolving needs.

"90% of decisions ignore 90% of the evidence" said Henderson, quoting a Gartner study. Big Data analysis can help providers test hypothesis and ensure they’re serving their community in the best way.

Want to learn more about how data is transforming care? Follow @MSUKINHEALTH on Twitter and get updates every day!