Proprietary, patented technology

Proprietary natural language processing technology
Optum’s CAC inpatient, outpatient, and professional applications feature proprietary LifeCode^® natural language processing (NLP) technology which deciphers the meaning and context of words in clinical documentation and recommends relevant ICD-9 and CPT® codes.

Defining NLP
NLP technology can read physician documentation, identify key clinical facts, and map those facts to codes. Physicians use standard dictation/transcription, speech recognition, or templates with free-text fields to record notes and then the NLP technology extracts information. Thus, the level of NLP software defines the success of the CAC solution.

The LifeCode advantage
LifeCode is the only patented NLP technology on the market today. LifeCode doesn't just search for word matches—it actually "understands" the meaning and context of the English language. The technology reads electronic documents and highlights all pertinent medical terminology to capture diagnosis and procedures that may have been overlooked in the traditional coding process. At present, Optum is the only vendor with a national, centralized engine. With this centralized approach, all ongoing updates and improvements to the engine are automatically received by all clients.

LifeCode:

• Contains over 10,000,000 medical facts
• Processes more than 5.5 million medical records each month
• Has coded more than 350 million patient encounters to date
• Features embedded NLP technology that can ‘read’ physician documentation, identify key clinical facts, and map those facts to codes
• Reads over 40,000 different physicians encounters nationwide
• Provides a central NLP engine for all customers

Dual patents, breakthrough technology
The LifeCode NLP engine combines the strengths of symbolic rules with statistical components—a patented NLP method. In fact, LifeCode is distinguished by two U.S. patents. The original patent—secured in 2005—describes “vector processing,” LifeCode’s mathematical model for isolating, comparing, and assigning different facts from clinical documentation to build a contextual framework.

In 2011, it received a second U.S. patent as the coding industry’s first NLP engine to accurately interpret the meaning and context of medical terminology in electronic health records (EHR). This patent describes “mere-parsing,” LifeCode’s method for determining meaning from free text. Parsing is defined as the syntactic analysis of words to determine grammatical structure. Mere-parsing is the process by which LifeCode assigns meaning not merely to single phrases within a sentence, but also to a combination of related phrases from throughout the documentation.

LifeCode has sophisticated inference rules that allow it to “understand” how documentation relates to coding rules, enabling it to, among other things, correctly assign combination codes, recognize related symptoms, or differentiate personal versus family history. These unique methods of organizing and extracting meaning from clinical documentation are the key ingredients that make LifeCode one of the most effective NLP technologies available today.

And, because LifeCode understands the words around such phrases, the software can make these distinctions, which results in more accurate coding recommendations and clinical analytics. These capabilities help increase the accuracy of the information to the insurer, increase the productivity of medical claims coding staff, decrease claims rejected by health plans or missed charges, and speed time to payment. NLP can also be used to improve the meaningful use of EHRs by identifying holes in current documentation practices and specific clinicians who may need additional support with regards to best practices.

CPT is a registered trademark of the American Medical Association.