A wide range of different technologies are poised to revolutionize the delivery of healthcare in the United States and beyond, from robotics to artificial intelligence. One technology that deserves some attention is known as natural language processing (NLP).
This technology makes it possible for computers to understand, interpret, and analyze spoken language as well as written text. development, NLP has been used in numerous ways, including personal voice assistants like Siri, language translation applications, and even spam email algorithms.
In recent years, there has been growing buzz about how NLP could improve the delivery of healthcare and make patient experiences more streamlined. Some of the key applications of NLP to healthcare include:
1. EMR dictation
Healthcare providers spend an incredible amount of time typing notes related to their patient encounters. These notes are essential for recording key information that other providers may need, as well as tracking progress and response to treatment.
Many providers rely on dictation technology to write notes, which is part of the benefit of NLP. Moreover, these dictation programs are becoming much more accurate. However, NLP can be used in much more exciting ways.
In the future, NLP could be used to transcribe automatically the conversation with patients while the appointment is happening. This means that providers spend much less time typing their notes and can therefore dedicate additional minutes to patient care.
The exciting part of NLP is its ability to add notes to an electronic medical record (EMR) in a structured way. NLP can automatically record and interpret these clinical notes based on the encounter. In addition, NLP can be used to analyze diagnostic reports to pull all key information into the EMR automatically.
2. Deep EMR analysis
An outgrowth of NLP’s ability to pull key information into an EMR is the possibility of analyzing these records to find key data points. For example, NLP could quickly scan the EMR of a patient to identify medications that have been formerly prescribed and indicate whether or not they were effective.
This data can save providers a great deal of time they would otherwise spend diving into the EMR. Researchers could also apply this technology to do deep dives into EMRs to answer important medical questions.
With NLP, it becomes much easier to analyze records based on geographic region or population segments since this entire process becomes automated. The administrative databases that now exist do not have nearly the same scale in terms of analyzing socio-cultural impacts on diseases and outcomes.
NLP has already been used in this way, such as with a 2016 study in Kawasaki disease that used an NLP-based algorithm to identify patients at risk. The overall sensitivity for the program was greater than 93 percent.
3. Review management
Providers and healthcare systems are both subject to online reviews, which can play an important role in such a regulated industry. Thousands of reviews can get posted each day, especially for large healthcare systems in big cities.
NLP can help manage this data. First, NLP can identify protected health information to ensure that it gets removed, as well as other content that is not compliant with HIPAA. Second, the technology can identify profanity or other content that could be offensive to people reading the reviews.
Third, NLP can help analyze the overall sentiment across a large number of reviews and the contexts of the comments made. This information can help healthcare systems improve their processes without dedicating an enormous amount of time to analyzing such a large amount of content.
Many healthcare systems are already using this technology, such as Temple University. NLP can also help track changes in sentiment to ensure that systematic changes had the intended effect.
4. Administrative streamlining
NLP has the potential to streamline many administrative processes in healthcare and save providers a significant amount of time. One of the biggest time-sinks for healthcare systems is securing payer prior authorizations. The complications involved with this process have only increased in the past few years.
Some companies are already working on NLP applications that can be used by a payer’s network to gain prior authorizations quickly. Moving forward, there is always the possibility that this type of system integrates with the EMR to determine when a prior authorization is necessary and automatically process it to avoid long wait times.
Another role that NLP can play is matching patients to clinical trials. A handful of companies are now working on systems that use NLP to match patients to clinical trials without the need for providers to do research and verify participation requirements on their own.
NLP could automatically match patients to these clinical trials to make the process seamless and allow them to make more information decisions about their healthcare. Moving forward, NLP could be used to streamline many other healthcare processes to save both patients and provides time.