Using Predictive Analytics To Slow Heroin Epidemic

Oct 3, 2017

The Hamilton County Heroin Coalition will use one quarter of a new $400,000 federal grant to predict who might be the next overdose victim and get them into treatment before it happens.

Quick Response Teams (QRT) arrive on the scene after an addict has overdosed and revived. They try to get the victim into treatment. Soon those teams will have information about the victim's friends in order to stop the next overdose.

Former Cincinnati Police Captain Dan Gerard, now with UC's Institute for Crime Science, says drug addicts are networked. Now part of the coalition, his team will give the data analytics to QRTs.

"There's a small number of places that are driving the overdoses and we also know that the addicts are networked. They are linked. They are trying to find who's got the best dope. They are sharing various techniques on how to acquire it," according to Gerard.

The QRTs will visit the first-degree friends in the network of the particular person who overdoses and let them know their friend overdosed and what treatment options are available.

Gerard says, "We can run it through some algorithms and take a look and identify both the people who are most likely to overdose and also more importantly some of the pro-social people as well."

The predictive analytics will start in the next several weeks.

Of the rest of the $400,000 DOJ grant, Hamilton County will designate $200,000 to programming and $100,000 to support staff.

The Coalition says, "In 2017 Hamilton County will spend $1.4 million to fund treatment and education programs and a 32-bed "recovery pod" at the Justice Center to reduce inmate relapse after release."