Updated March 4, 2022
Notice ID: 36C24E22Q0079
Related Notice: 36C24E22Q0079
“The objective of this pilot project is to execute a Quality Improvement (QI) study to evaluate the value of machine learning (ML) to decrease the likelihood of recurrence of opioid use among Veterans. The findings from this study will provide the foundation for a subsequent training or MERIT grant proposal to the VA with the goal of using ML within a decision-support tool to help decrease the likelihood of recurrence of opioid use among Veterans.””
“Scope of Work: The project will apply state-of-the-art statistical analysis on Veterans Affairs Pittsburgh Healthcare System (VAPHS) Computerized Patient Record System (CPRS) data to develop precision statistical models for Medication Assisted Therapy (MAT) selection for Veterans with Opioid Use Disorders (OUD). These models will incorporate patient characteristics, medical history, and treatment regimens to predict risk of recurrence of opioid use for individual patients after treatment for OUD. Data will be accessed by the VA Investigator through the VA Informatics and Computing Infrastructure (VINCI) portal…”
“Key Tasks (Primary Responsible Entity)
- Extract and clean CPRS data for Veterans served by VAPHS who are diagnosed with OUD. (VA)
- Apply ML algorithms to develop models to predict recurrence of use in OUD patients who have received treatment for OUD. (Contractor)
- Write technical report summarizing model performance and outlining potential next steps to utilize ML results to decrease recurrence of opioid use. (Contractor)”
Posted February 18, 2022
Notice ID: 36C24E22Q0079
“The Department of Veterans Affairs, Office of Research and Development, intends to contract on a sole source basis with SOAR Analytics, LLC to provide advanced statistical analysis services to Veterans Affairs Pittsburgh Healthcare System (VAPHS). SOAR Analytics, LLC will used Computerized Patient Record System (CPRS) data to develop precision statistical models for Medication Assisted Therapy (MAT) selection for Veterans with Opioid Use Disorders (OUD). These models will incorporate patient characteristics, medical history, and treatment regimens to predict risk of recurrence of opioid use for individual patients after treatment for OUD.”