The Natural Language Clinical Information Solution
The fully-developed natural language medical information and coding system presents the immediate opportunity to create a centralized, Internet-accessible, HIPAA compliant universally accessible source for all clinical information from all hospital admissions in the United States. The application operates as an Internet-based system whereby a hospital electronically submits all clinical information that is in text form in every medical record. The application captures all clinical information in its database and performs an analysis of more than 19,000 ICD-9CM codes to identify all such codes that are supported by the substantive meaning of the text. Both the ICD codes and the clinical data are returned to the submitting hospital where the coder elects which of the identified codes will be included to create the DRG for billing purposes.
Each hospital that uses the application will have permanent access to its own natural language database and will be able not only to recreate all ICD-9CM codes that the system identified, whether they were used in billing or not, but also to mine all of its own data free of limitations imposed by other database systems that are restricted by predetermined data entry parameters. Abstraction of medical information is not limited to ICD nomenclature, and customized rules can be incorporated to analyze any number of variables. Because the application is a stand alone Internet-based application, it stands outside of the hospital's existing information network and merely receives information in text format. The time required to become fully operational usually will be less than thirty days, and the start-up costs will be nominal.
The ultimate end user for this technology, however, will not be individual hospitals or hospital systems. The ability to compile a natural language database from the information contained in the most commonly-used dictated medical records (History and Physical, Consult Reports, Operative Reports, Discharge Summaries) will enable government agencies such as CMS, the VA, the CDC, and the NIH to create a database of all clinical information from all hospital admissions in the United States, including real time information for patients undergoing current treatment. Those agencies will be able not only to utilize that information to carry out their own governmental functions relative to healthcare management and medical research but also to make all such information, absent patient identifiers, available to researchers, pharmaceutical companies, healthcare product manufacturers, and all others who might benefit from access to an unlimited source of clinical information. If implemented by the major governmental payers, within five years the application could be used to create a National Clinical Information Database that could be queried in a manner similar to existing commercial sites, but with a natural language search engine that is far more specific.
InferRx will enable hospitals, payers, government agencies, and researchers to convert to ICD-10, and to analyze both past and future clinical information using ICD-10, with no disruption and at greatly-reduced costs.