Sustaining Recovery for People on Opioid Agonist Treatment With Conversational Agents

Engaged in buprenorphine treatment or treatment retention is defined as a participant's treatment period of at least 365 days, including those with multiple disengagement and reengagement periods, as long as any gap in active prescriptions or receipt of injections is less than 60 days.

Treatment retention is defined as a participant's treatment period of at least 365 days, including those with multiple disengagement and reengagement periods, as long as any gap in active prescriptions or receipt of injections is less than 60 days.

Treatment duration is defined as length of continuous buprenorphine treatment in days until first discontinuation and will be measured via electronic health record (EHR) prescription and injection data and supplemented with self-report. Discontinued buprenorphine treatment occurs once the EHR indicates that they haven't filled a buprenorphine prescription or haven't received a buprenorphine injection in the previous 60 days. The treatment discontinuation date will be the last day of an active prescription or injection received.

Opioid use will be measured by self-report via the modified Addiction Severity Index (ASI). The modified ASI is administered by the study staff for a list of 20 drugs/substances. For each the number of days of use and the route of administration of that drug/substance [0=N/A, 1=oral, 2=nasal, 3=smoking, 4=non-intravenous (IV), injection 5=IV, or 8=refused] is documented.

The SU-SMS is a self reported, 18-item instrument. Each item has a 5 Likert frequency responses from 1=Never to 5=Very often. Total scores can range from 1 to 90 or from 1-30 for each of the three types of stigma [anticipated, enacted, and internalized]. Lower scores are correlated with less stigma.

BT-SMS is a self reported, 25-item instrument. Each item has a 5 Likert frequency responses from 1=Never to 5=Very often. Total scores can range from 1 to 125 or from 1-45 for anticipated stigma, 1-45 for enacted stigma, and 1-35 for internalized stigma. Lower scores are correlated with less stigma.

Changes in recovery capital will be measured by self-report via the Brief Assessment of Recovery Capital (BARC-10) which is a 10-item self administered instrument. Each item has 6 Likert possible responses where 1=Strongly disagree to 6=Strongly agree. Total scores can range from 1 to 60. Higher scores are correlated with more recovery capital.

Frequency of interactions with the ECA app will be collected via the app itself and provided to the BMC study team by the investigators at Northeastern University.

User attitudes and satisfaction with the ECA app will be measured by self-report using the Bond subscale of the Working Alliance Inventory which is an investigator developed instrument. The Bond subscale has 6 questions with an analog type response from 'disagree completely' to 'agree completely'.