Risk Prediction Projects

Overdose Risk Prediction Projects

Traditional methods for determining risk of opioid overdose don't take into account the complex interactions among risk factors to determine a person's individual risk. We use machine learning to handle complex interactions in large data sets, discover hidden patterns, and generate actionable predictions for clinical care settings.

Machine-Learning Prediction and Reducing Overdoses with EHR Nudges (mPROVEN)
PI: Walid Gellad, MD, MPH
Funding Source: NIH/NIDA; July 2022 - April 2027

PDMP Overdose Data to Action (OD2A) Opioid Overdose Surveillance
PI: Walid Gellad, MD, MPH
Funding Source: CDC; March 2020 - August 2023

Machine learning and opioid overdoses in Allegheny County
PI: Walid Gellad, MD, MPH
Funding Source: R.K. Mellon Foundation; July 2018 - June 2020, July 2020-December 2023

Substance Use Treatment Projects

We conduct research to inform access to and development of effective substance use treatment, especially among Medicaid enrollees.

Geographic Access to Medication Assisted Treatment for Medicaid Enrollees with Opioid Use Disorder
PI: Coleman Drake, PhD
Funding Source: NIH/NIDA; July 2020 - June 2024
Funding Source: National Center for Advancing Translational Sciences; June 2019 - May 2021

All published work in opioids and risk prediction