Psychatory © Copyright 2018. All rights reserved.

Artificial Intelligence Tool Encourages Peer Group To Address Substance Abuse

For several years, statistical data have shown consistent increase in the number of people who become addicted to illegal substances. Based on the 2015 data from the Centers for Disease Control and Prevention, the prevalence rate of individuals aged 12 years and above who use illicit drugs reached 10.1 percent.

Data from the National Institute on Drug Abuse (NIDA) via their 2014-2017 report titled “Monitoring the Future Study: Trends in Prevalence of Various Drugs” shows an increase in use among 8th graders, 10th graders, and 12th graders in almost all types of substances.

Positive Peer Groups Encourage Recovery

Previous research has established that peer groups have an important role in curbing substance abuse, particularly when it comes to the recovery phase. Groups that engage in intervention can influence recovery and relapse. While exposure to negative behaviors may push at-risk individuals to go into relapse, positive peer group may also help in making them stick to their recovery program.

A team of researchers from the University of Southern California Center for Artificial Intelligence in Society devised an algorithm that could help in grouping participants of intervention programs. The researchers worked with a Denver-based non-profit for homeless youth, Urban Peak, to develop the program.

By sorting participants into smaller groups, they become more involved and focused on maintaining healthy relationships and veer away from negative social connections. Researchers theorized that if participants are improperly grouped, intervention programs may have the tendency to worsen the problem and could result to “deviancy training,” wherein peers try to reinforce each other’s bad behavior.

In formulating the artificial intelligence (AI), the researchers created a program that would sort individuals into groups based on their social ties and previous history of substance abuse. Results of their study showed that the algorithm could sort intervention group participants much better than utilizing control methods.

They identified that forming uniform subgroups of regular substance users is not always the best model for group intervention.

“Uniform distribution of users while ignoring their existing relationships can greatly decrease the success rate of these interventions,” said lead author and USC computer science graduate student Aida Rahmattalabi, according to the USC Viterbi website. “In some cases, we found it’s actually a bad idea to conduct the intervention: for example, if you have many high-risk people in a group, it is better to not connect them with low-risk individuals.”

Results of the study (PDF), titled “Influence Maximization for Social Network Based Substance Abuse Prevention,” are published in the student abstract section of the AAAI conference on Artificial Intelligence.

Having an effective intervention group therapy is crucial in keeping recovery in check. Of the millions of individuals who abuse substances, only about 1 in 10 attends group therapies. Those who attend are not really successful in curbing their addiction. This is partly due to the ineffectivity of the approach used.

In the report published by the National Center on Addiction on Substance Abuse, it was stated that many of the people needing addiction treatment do not have access to evidence-based care. Many of the intervention programs used are not aligned with what is known to effectively work.

Write a response

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.