A new prediction model is being developed by scientists in response to rising suicide rates among members of the military. The program will use an algorithm to screen soldiers and flag those whose information contains multiple factors that align with those of an individual who is at a high risk for suicide. Doctors could then use this information to take proactive, preventative measures to reduce the risk of these patients killing themselves.
The system itself is a computer program that retrieves information on over 20 factors associated with an individual and rates them to determine their level of risk. Factors include things like prescription drug use, previous psychological conditions, age at enlistment and criminal offenses. There are some ethical concerns associated with this program due to the fact that the information is mined directly from the individual’s military medical history. Some believe that this is an invasion of privacy and could hinder the careers of military members who were found to be at a higher risk.
Scientists are developing this new prediction model in response to suicide rates among soldiers which have been rising since 2004. Historically, the rate of suicide in U.S. army soldiers has been below that of civilians. Since 2004, after the start of the conflict in Iraq, the suicide rate in the Army has risen above that of the civilian population. The model will help doctors predict which patients are at the highest risk, giving them an advantage in treating them.
The prediction model is based on a study performed by Ronald C. Kessler, who is a health care policy professor at Harvard, and his research team. Kessler wanted build an “actuarial risk algorithm” with the intention of using it to provide better treatment to soldiers. According to the New York Times, Kessler and his team analyzed 40,820 soldiers’ records who had been placed in a hospital for a mental health problem at least once from 2004 through 2009. Patients who have been hospitalized are known to be at risk months after being discharged, the researches set forth to discover what factors the ones who actually committed suicide shared.
The researchers began by making a list of over 300 different factors that might be related to the level of risk. These included things like military rank, substance abuse and access to weapons. After testing the list over and over the researchers reduced the number down to select group that when seen together identified the highest 5 percent who were at risk. This 5 percent accounts for over half of the individuals who committed suicide within a year of leaving the hospital.
Using this new prediction model to discover which military members are at a high risk of suicide, doctors would be better able to treat those soldiers using preventative techniques that would be centered around their specific needs. Examples of these measures would include things like required outpatient therapy after discharge, working with the individual to build a support group via their family and friends and teaching them coping mechanisms for them to use when they begin to feel overwhelmed.
By Clara Goode
Photo Source: Robb & Jessie Stankey – Flickr License