VA researching natural language processing in EHRs to prevent suicide

The U.S. Department of Veterans affairs is researching the use of natural language processing within its electronic health records system to automate suicide risk alerts.

According to an announcement from the VA, natural language processing is "part of the technology that makes Google work," and ongoing research at the VA Puget Sound Health System and the University of Washington is seeking to learn how the VA can use it.

VA research has found that past suicide attempts are the most compelling sign of future risk--even stronger than major depression; their studies show that a past suicide attempt raises the odds of going through with suicide by 40 times.

The researchers used data on more than 100,000 veterans to query text from doctor's notes, trying to hone in on "red flags" indicating past suicide attempts. They developed "an automated text search that was about 80 percent accurate, compared against manual checking of each record by a psychiatrist," according to the announcement.

"The electronic medical record system stores a very large body of clinical notes," Ken Hammond, M.D., a former VA psychiatrist and current researcher, said in the announcement. "We've shown that we can use search engine technology to more easily identify those Veterans who have attempted suicide at some point in their lives. That can help us prevent future attempts."

The issue of suicide among veterans reached its highest point in 10 years in 2012. That year, the VA expanded telehealth for mental health services and removed patient co-pays. In late July, it was reported that 80,000 vets used telehealth services in more than 200,000 remote consultations in 2012.

To learn more:
- read the VA announcement

Related Articles:
VA launches two mHealth pilot programs
7 health technologies developed for veterans
VA grants for women's health programs expand telehealth efforts
VA blazing its own trail when it comes to government mHealth