An Artificial Intelligence-Based Smartphone App for Assessing the Risk of Opioid Misuse in Working Populations Using...

Area of Science:

• Mobile Health (mHealth)
• Artificial Intelligence (AI) in Healthcare
• Occupational Health

Background:

• Opioid Use Disorder (OUD) is a significant crisis in the US, affecting over 10 million individuals.
• Physically demanding industries like transportation, construction, and healthcare have high rates of OUD.
• OUD leads to increased costs, absenteeism, and reduced productivity in the workforce.

Purpose of the Study:

• To develop a smartphone application for tracking work-related risk factors associated with OUD.
• To focus on high-risk occupational groups susceptible to OUD.
• To utilize mobile health tools for interventions outside clinical settings.

Main Methods:

• Developed a smartphone app through literature review and expert panel evaluation of risk assessment questions.
• Utilized synthetic data for training and testing, avoiding human participant data.
• Employed a Naive Bayes machine learning algorithm to predict OUD risk.

Main Results:

• The developed smartphone app is functional and successfully predicted OUD risk using synthetic data.
• The Naive Bayes algorithm demonstrated efficacy in risk prediction.
• The study established a platform for future testing with human participant data.

Conclusions:

• Mobile health apps show promise for disease prediction and prevention, including OUD.
• The app ensures user privacy and accuracy through REST API and cloud encryption.
• Offers tailored mitigation strategies for high-risk workforces, potentially aiding in reducing the opioid crisis.