Until recently, the use of natural language processing (NLP) in healthcare has been primarily limited to research efforts and population health within academic medical centers. However, with the proliferation of unstructured data from electronic medical records, providers are now seeking to harness the potential of their data and considering a variety of use cases for NLP technology. That’s the conclusion of a recent KLAS report entitled “Natural Language Processing: Glimpses into the Future of Unstructured Data Mining.”
The report includes insights from 58 provider organizations and examines the various ways providers are currently leveraging NLP technology, as well as some of use cases poised for wider adoption. Coding and documentation applications represent the broadest use of NLP engines. But it is clear providers have a growing interest in NLP solutions that advance their population health initiatives. An increasingly popular use case, involves applications that use NLP to mine unstructured data within patient populations and include predictive analytics to identify at-risk patient populations.
A few of the major findings from KLAS’s report are summarized below.
How is NLP being used today?
About 36% of the surveyed providers were considered “preliminary” users of NLP for applications such as computer-assisted coding (CAC), clinical documentation improvement (CDI), patient-summary applications and/or a single use case. Thirty-one percent of the organizations were “developing” users that leveraged NLP for computer-assisted physician documentation (CAPD), patient summaries, population health, or research. Twenty-five percent of the providers were still “considering” NLP and not yet live with any applications, while 8% were “established” NLP users with mostly homegrown solutions designed for research and population health use cases.
Research and population health use cases
Established users of NLP technology, as well as many developing users, are relying on NLP for research or population health initiatives, including research cohort identification; at-risk populations and risk adjustment; decision support; and regulatory, quality, and safety reporting. One of the earliest use cases for NLP was for the mining of data to identify possible subjects for clinical research trials. NLP’s data-mining capabilities can also help providers identify at-risk patients within a population so that risk can be assessed under value-based care models. According to KLAS, the identification of at-risk patients is the use case that providers mentioned most often when asked about future plans or interests for NLP. KLAS verified that Linguamatics have active customers in the population health space.
Providers are also looking to NLP technology to address decision support needs. The NLP engine can identify key information about a population or patient, intelligently filter the data, and then advise on the appropriateness of actions taken or tests ordered. KLAS identifies regulatory, quality, and safety reporting as a final use case. Using NLP technology, providers are able to extract critical quality and patient safety information from unstructured data in order to address their compliance reporting needs. KLAS highlight one Linguamatics customer in particular who is using I2E for predicting risk of 30-day readmission.
Coding and documentation use cases
KLAS found that most preliminary or developing users of NLP technology were relying on vendor-driven software (as opposed to homegrown solutions) to address four specific use cases: CAC/CDI, CAPD, code reconciliation, and patient summaries/patient documentation reconciliation.
With CAC, the software suggests codes based on documentation to increase charge capture, while CDI solutions provide workflows to improve the quality and accuracy of clinical documentation. The use case for CAPD is similar and has the potential to improve the accuracy of documentation and decrease clinician error. With CAPD, the application relies on NLP to provide physicians with real-time bedside feedback on charting and coding, and suggests diagnoses and terms related to the current symptoms and diagnosis.
Coding reconciliation applications use NLP to review documentation and ensure providers are billing for all possible codes. Similar text-mining technologies can also be used to develop patient summaries and patient documentation reconciliation and Linguamatics is again noted in this area.
Linguamatics and NLP
Customers currently use Linguamatics’ NLP-based solutions for a variety of use cases, including patient summary creation and the identification of at-risk populations. Our clients are using NLP to support cohort selection across populations based on unstructured data, as well as to improve clinical documentation and information extraction workflows.
Some of the key growth areas we see include support of precision medicine initiatives, risk analysis, and managing population health, as well as addressing the regulatory burdens of tracking, measuring, and reporting on quality. As a provider of agile, real-time, scalable NLP-based text mining technology, we believe in the power of NLP and recognize its promise for addressing the challenge of transforming large volumes of unstructured patient text into actionable insights and enhanced knowledge.
For more information register for the upcoming HealthSystemCIO.com webinar KLAS Reports: Natural Language Processing — Glimpses Into The Future Of Unstructured Data Mining on August 23rd.
 KLAS report, “Natural Language Processing: Glimpses into the Future of Unstructured Data Mining,” April, 2016, 8-9.
 Ibid, 11.