Today, let’s talk about talking. You might be surprised that talking is
going hi-tech, but it is! For years, medical care providers have been dictating
medical records and transcription
companies have been typing those records. Innovations in speech recognition developed
wonderful products like Dragon Naturally Speaking that changed how many
providers document medical care. Now, speech recognition has gone to new levels
with Natural Language Processing (NLP). NLP is not really new. It has been in
development for well over 20 years but now it is catching on and changing
healthcare as we know it.
So, what is NLP? Natural Language Processing is a form of artificial
intelligence that, when used in healthcare, can identify important key
components in a medical record and transform a narrative text (e.g. a story)
into coded discrete data elements output in an XML format. With that said in
such fancy terms, I will translate it to plain English with an example.
Let’s say that a doctor dictates the following statement: “John Doe is
a 59 year old male with type 2 diabetes.”
NLP technology can take that statement and identify each important item
in the statement individually, give it a header or a tag, and code it (if
applicable), a little like this:
patient name: john doe
age: 59
gender: male
diagnosis: type 2 diabetes
ICD-9 diagnosis code: 250.00
This is some impressive stuff! It opens up a world of automation
related to regular-old speech in medical practices and hospitals. Innovative
transcription companies, like StenTel, for example, are working with NLP
technology from NLP developers such as MedLEE, CodeRyte, and IBM. Other companies
like 3M, and Nuance (the makers of
Dragon Naturally Speaking) are also using NLP technology in a variety of ways
from highly automated encoders to clinical documentation improvement to
population health management.
For these technology companies and the providers of
NLP technology, goal is two-fold. First, NLP can help improve provider
reimbursement by helping coders to accurately identify patient conditions in
otherwise messy and at times voluminous medical records. Secondly and perhaps
more importantly, this technology opens the door for data mining and
aggregation in ways that were never possible before. This is the heart and soul
of the population health movement.
With NLP technology, automated systems are able
to warn us about impending outbreaks and so-called ‘clusters’ before they get
too big. When we know where these condition ‘clusters’ are (cancer clusters,
autism clusters, etc.) and when they are beginning, we are better equipped to find
for the cause and implement the appropriate precautions to prevent further
problems. Additionally, NLP offers the ability to analyze outcomes to see what
treatments work and what treatments do not work for various conditions and
diseases. With this kind of information, healthcare (diagnosis and treatment)
will improve and become more affordable – all thanks to the handful of
innovators developing and promoting NLP.
