Question normal

EHRs today often rely on structured data input formats (such as check boxes or drop down menus) in an attempt to limit data inconsistencies. These formats are cumbersome, frustrating, and take the joy out of medicine by removing the narrative structure of note writing. A better system would incorporate natural language processing technology such as Google’s Hummingbird in order to allow free text, story-like inputs by physicians. The EHR would then package this data using common terminology such as ICD10 or SNOMED for billing as well as to maintain a universal problem list and medication history. This data would then be stored separately in a universal database, rather than intertwined with each individual software platform. This would allow easy communication between systems. At the same time, it would be easier to customize the user interface and apply new innovations such as the TwinList medication reconciliation.