When it comes to clinical workflows, there are a variety of ways providers enter information into electronic health records (EHRs). One frequently leveraged part of the EHR is the notes section – where clinicians can jot free-form text descriptions of a patient’s symptoms and treatment.
But while providers often take advantage of this documentation method due to its ease of use, it comes at a cost. These notes aren’t always easy to standardize, which is critical for other needs like billing and reimbursement. And the unique, nuanced language of healthcare providers makes standardization a challenging task – even when tools like natural language processing (NLP) are used.
So, what types of notes might an NLP solution be tasked with “reading” in healthcare? Here’s an example:
A surgical note is often full of abbreviations and acronyms. However, a well-trained NLP solution should be able to read the above as: “Patient returns, postoperative day 10 for routine follow-up status post right hemicolectomy. Mild abdominal pain, tolerating clear fluids, endorses mild right upper quadrant pain. Wound clean and dry, CT of abdomen scheduled in the morning.”
The acronyms, abbreviations, and misspellings seen in this example are just the tip of the iceberg when it comes to distinctive language in clinical documentation. Download our latest insight brief, Reading like a human: The key to successful natural language processing in healthcare, to learn about other situations that an NLP solution will likely encounter when extracting information from clinical documentation in free-form text.