The digitization of health care has transformed how data is collected and managed, creating a deluge of information that includes everything from EHRs to clinical data to payer and provider statistics.
Natural language processing (NLP) is starting to deliver capabilities that solves for some of the complexities seen in the payer ecosystem.
NLP provides a new and innovative way to connect the dots – to look at a wide range of information and find little nuggets of data that might have gone unrecognized, or identify a powerful correlation that hadn’t been seen before. Handwritten notes, dictated patient interactions, blood test results, X-rays, labs– they all become part of a powerful data picture than can be exploited. NLP enables new insights across a range of settings – from looking at the effectiveness of treatment plans to measuring myriad patient outcomes to identifying common factors for specific diseases.
NLP leverages cognitive algorithms which allow a computer to “read” unstructured text and pick out key words and phrases that can then be analyzed to understand their meaning in a specific context – in this case healthcare. This allows caregivers to tap into the vast, previously unexplored reams of data that are simply unreadable by standard tools designed only to provide insight within a set of specific data.
To date, most of the analysis of unstructured health care data has been manual, resource-intensive, and error-prone. Until recently, the bulk of this kind of work was limited to the relatively small amount of discrete data (approximately 20%) found in the EMR, which are typically entered in unique formats, such as the required selection of specific ranges of numbers for vitals or laboratory data fields.
But with the increasing adoption of natural language processing along with Big Data and machine learning, this approach to analysis will be transformed. And as patient and payer data volumes continue to grow exponentially, traditional methods will simply no longer be viable.
“We’re finding new, different types of dictated documents we haven’t used,” says Scott Evans, director of medical informatics at Intermountain Healthcare in Salt Lake City, Utah. “Since the amount of dictated documents and unstructured data is growing, the need for NLP in healthcare is also growing.”
Based on current uses, the next step for NLP will be the increased use of machine learning to expand the power of this approach. Soon, users will be able to not just put healthcare information in a broader more strategic context, but they will also be able to exploit new levels of predictive analytics and artificial intelligence to help clinicians in the decision-making process.
Healthcare organizations that take advantage of natural language processing and other data-driven technologies will successfully manage the transition to value-based care and ultimately be positioned to achieve superior clinical, financial, and operational outcomes.