Overview
Chronic pain is a prevalent and debilitating condition among Veterans, often accompanied by depression, anxiety, sleep disturbances, and social isolation. Despite the availability of behavioral interventions many Veterans-particularly those with high in-person healthcare utilization-face barriers to accessing non-pharmacological pain management strategies. These barriers include limited clinician availability, geographic constraints, and digital exclusion among older or underserved populations.
This study aims to evaluate the feasibility, acceptability, and preliminary impact of My AI Companion, an automated, text/telephone-delivered chatbot designed to support chronic pain self-management among high-utilizing Veterans. The intervention offers a low-barrier approach to promoting symptom tracking, patient activation, and engagement in coping strategies-without requiring Internet access or clinician time.
Description
This is a prospective cohort study with a retrospective matched comparison group. We will enroll 200 Veterans receiving care at VA Augusta Health Care System who meet criteria for high in-person healthcare utilization and have a diagnosis of chronic pain. One hundred participants will use the chatbot over a 12-month period and complete assessments at baseline, 3, 6, and 12 months. Primary outcomes include pain interference (PEG Scale), quality of life (VR-12), patient activation (PAM), self-efficacy, depression and anxiety symptoms (PROMIS), sleep disturbance (PROMIS), loneliness (UCLA), and substance use (AUDIT-C, DAST-10). We will also collect data on employment, caregiving roles, and opioid prescriptions. A retrospective comparison group will be identified from medical records to assess relative changes in healthcare utilization. The remaining 100 participants will act as the comparison group in this study is primarily identified through retrospective chart data, we will proactively invite a subset of these Veterans to complete the same self-reported measures as the intervention group (e.g., PEG, VR-12, PROMIS Depression/Anxiety, Sleep, Self-Efficacy, UCLA Loneliness, AUDIT-C, and DAST- 10). This approach will strengthen our ability to compare patient-reported outcomes across groups and provide valuable insights into the impact of the chatbot intervention. We will identify Veterans in the comparison group based on matched characteristics (e.g., high in-person healthcare utilization, chronic pain diagnosis, and demographic/clinical factors) and invite them to participate in survey assessments at baseline, 3, 6, and 12 months. Outreach will include an initial letter or email invitation explaining the study purpose, eligibility criteria, and procedures, followed by phone calls from the research team to answer questions, obtain consent, and support survey completion.
Eligibility
Inclusion Criteria:Participants must meet all of the following eligibility criteria to be enrolled in the study:
- Be a Veteran receiving care at the VA Augusta Health Care System Pain Medicine Service.
- Be aged 18 or older.
- Have a documented diagnosis of chronic pain (defined as pain persisting longer than 3 months).
- Be classified as a high utilizer of VA in-person outpatient services, defined as having 4 or more in-person primary care, behavioral health, or pain-related visits within the past 6 months.
- Be English-speaking.
- Have access to a telephone (landline or cell phone).
- Have capacity to provide informed consent.
Exclusion Criteria:
- Individuals will be excluded from the study if they:
- Have moderate to severe cognitive impairment that would interfere with the ability to engage with a text/telephone-based chatbot (as determined through medical record review or clinical judgment).
- Have active psychosis or acute psychiatric instability (e.g., recent psychiatric hospitalization in the past month) that would impair study participation.
- Are enrolled in another conflicting behavioral intervention study.
- Are non-English speaking, due to the current availability of the chatbot only in English.
