Q&A: Unstructured Clinical Notes Just As Valuable For Analytics

While much has been made of the need to standardize clinical data with EHRs that use templates, click boxes, and dropdowns to funnel information into pre-determined data elements, the need for such strict organization may not be top priority for long.  Advances in natural language processing of clinical documentation mean that physicians can say what they want without excluding narrative text from analytics.  Juergen Fritsch, Chief Scientist at M*Modal, sat down with HealthITAnalytics to discuss how to extract the value from unstructured data in the pursuit of better patient care.

What are some of the challenges when using clinical documentation for analytics?

Traditionally, people have relied on using structured documentation to enable analytics.  They’ve designed their documentation forms to capture the data that they wanted for analytics.  So it started with patient demographics and it went from there to medications and allergies and things like that, so they could then use that structured data to analyze the patient population.