According to the Department of Veterans Affairs publication “Research Currents”, the VA is working on a new research project to harness the power of Natural Language Processing (NLP) to obtain more value from electronic records. The project called the “Consortium for Healthcare Informatics Research” (CHIR) is working to change free text as found in the doctor’s notes in electronic medical records and then translate these notes into structured data. It’s basically a matter of converting information from one form of information into a different format.
Webmasters and database architects typically use checkboxes, pull-down menus, and other tools to structure the information they want to collect. Electronic health records use structured data and templates wherever possible but there is also the need for free text.
It is difficult to ask clinicians to input only structured data when they are evaluating patients as they need to be able to use free text. Merry Ward, PhD, who oversees CHIR for the VA’s Office of Research and Development, says “narrative is very important for healthcare providers. Radio buttons, pull-downs, yes-no, and other forced choices can only go so far in describing the patient’s condition.
According to the researchers, it is also important to be able to convert narrative text into structured data, because it makes it easier to measure quality, track infectious diseases and adverse drug events, and develop new decision support systems.
As part of CHIR, two multi-year studies are underway to develop natural language processing and clinical applications. One study will look at MRSA to be able to capture more relevant MRSA information in the record beyond the standard coded data. A second project on PTSD will examine whether free text can shed light on how the disorder progresses and how symptoms may vary from one patient to the next. Other CHIR projects will look at clinical topics that deal largely in caring for heart disease, cancer, and other diseases.
In another project, the “Veterans Informatics, Information and Computing Infrastructure” (VINCI) will roll out electronic medical records from VA sites nationwide into one secure centralized data repository. Tools to do natural language processing will be installed and available for use within the VINCI environment to enable access to not only free text but also to have access to information on all veterans. Researchers will then be able to conduct large, nationwide studies covering up to two decades of medical care for patients from the information documented in the electronic records.