Towards Data-Driven Artificial Intelligence Models for Monitoring, Modelling and Predicting Illicit Substance Use

Illicit substance use (ISU) is a major public health problem and a significant cause of morbidity and mortality globally. Early assessment of risk behaviour, predicting, identifying risk factors, and detecting illicit substance use become imperative to reduce the burden. Unfortunately, current digital tools for early detection and modelling ISU are largely ineffective and sometimes inaccessible. Data-driven artificial intelligence (AI) models can assist in alleviating the burden and tackling illicit substance use but their adoption and use remain nascent. This study applied the PRISMA model to conduct a systematic literature review on the application of artificial intelligence models to tackle illicit substance use. The study revealed that elastic net, artificial neural networks support vector machines, random forest, logistic regression, KNN, decision trees and deep learning models have been used to predict illicit substance use. These models were applied to tackle different substance classes, including alcohol, cannabis, hallucinogens, tobacco, opioids, sedatives, and hypnotics among others. The models were trained and tested using various substance use data from social media platforms and risk factors such as socioeconomic and demographic data, behavioural, phenotypic characteristics, and psychopathology data. Understanding the impact of these risk factors can assist policymakers and health workers in effective screening, assessing risk behaviours and, most importantly, predicting illicit substance use. Using AI models and risk factors to develop data-driven intelligent applications for monitoring, modelling, and predicting illicit substance use can expedite the early implementation of interventions to reduce the associated adverse consequences.