Case Study: Leveraging natural Language Processing to Improve Quality of Care and Reduce Costs

October 23, 12:00am, MDT - 12:00am, MDT

Registration

Non-member Price: 
$0.00

Case Study: Leveraging natural Language Processing to Improve Quality of Care and Reduce Costs 
by R. Scott Evans, PhD, FACMI


Dr. Evans addressed a group of UHIMSS members, giving us a taste of his upcoming publication on using NLP to improve quality of care. Given the problem of deep venous thrombosis (DVT) developing in patients who had peripherally inserted central catheters, Dr. Evans began with the philosophy of “You can only manage what you measure”: he set about gathering patient data. He quickly realized that most of the needed data was to be found in physician and nursing notes, with the inherent problems of each group using slightly different verbiage – and spellings – and not necessarily in neat, fielded data. Dr. Evans found that a solution to translating and organizing that written data in a homogeneous fashion needed to occur prior to having the ability to analyze this data to find clues to the problem. Evans walked us through the NLP data translation, data analysis and subsequent refining of both the processes and the needed data. He also shared with us some of the results, which have resulted in changes to patient care protocols and a sizeable reduction in DVTs. This project has been so successful that it is now being expanded to identify DVTs in a broader patient population.

Unfortunately, Dr. Evans is unable to share his presentation slides with UHIMSS at this time, as his results are about to be published. Many of Evans’ papers can be found online.