Healthcare AI Innovation
A Pilot Study of an AI Chatbot for the Screening of Substance Use Disorder in a Healthcare Setting
This pilot study evaluated the usability and potential efficacy of "Be Well Buddy," an AI chatbot designed for Substance Use Disorder (SUD) screening and referral in a healthcare setting. Implemented through Be Well Texas, the chatbot provided information on SUD, anxiety, depression, and treatment options, facilitating self-screening over a seven-day period. Of 92 participants, 91 engaged, and 32% completed at least one screener. Key findings include high user comfort with screening, positive user feedback on helpfulness and non-judgmental tone, and a 50% appointment rate among referred individuals, demonstrating its potential to increase SUD screening rates beyond traditional methods (typically <10%). Despite minor technical errors, the system proved memorable, efficient, and acceptable, suggesting AI chatbots can significantly improve access to crucial SUD services.
Key Impact Metrics
0
Participants Completed Screening
0
Screened Individuals Referred for Care
0
Referred Individuals Made Appointments
0
AI Chatbot Response Precision
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
AI Chatbots for SUD Screening
This study introduces "Be Well Buddy," a novel AI chatbot specifically designed for Substance Use Disorder (SUD) screening and referral. Unlike generic generative AI, this chatbot leverages a curated closed library of expert-created content, ensuring accuracy, non-stigmatizing language, and theory-based health communication strategies. This approach directly addresses the ongoing opioid crisis in the U.S. by providing an accessible, anonymous, and non-judgmental pathway for individuals to gain information, self-screen for SUD, anxiety, and depression, and receive direct referrals for treatment. The system’s potential to significantly increase screening rates (traditionally less than 10%) highlights its critical role in expanding access to care.
32%
of participants completed at least one screener, significantly higher than the national average for at-risk individuals (<10%).
Usability & User Acceptance
User feedback confirmed high levels of satisfaction and ease of use for the Be Well Buddy chatbot. Participants found the system "unobtrusive, convenient, and easy to use," appreciating its informative and non-judgmental tone. The supportive language helped users feel comfortable self-screening, a crucial factor for sensitive topics like SUD. The system's prompts for further engagement were also valued, preventing disengagement. While most users (9 of 13 interviewees) utilized the self-screening tool, some expressed a desire for direct connection with live support for complex questions or location-specific resources, highlighting areas for future enhancement.
Enterprise Process Flow: Be Well Buddy Engagement
Clinic Chat invites Participants to chat with Be Well Buddy via SMS
→
Participants engage with Be Well Buddy, receiving periodic screening invitations
→
Participants self-select to complete screeners (SUD, anxiety, depression)
→
If high score, referred to Be Well Texas for appointment; If neutral, continue chatting & encouraged to screen later.
Performance & Efficacy Potential
The Be Well Buddy system demonstrated strong technical performance, correctly responding to 80% of user-initiated queries. While minor delivery errors occurred (6% of messages impacted over 9% of delivery days), these are deemed addressable in future iterations. Critically, the system exhibited significant potential for efficacy in increasing access to care: 32% of enrolled participants completed at least one screener, with 83% of those screened receiving a referral for treatment, and half (50%) of those referred subsequently making an appointment. This outcome is particularly promising when compared to the <10% screening rate observed nationally for at-risk individuals, and shows improved completion rates over traditional website screening (3% abandonment rate vs. 39%).
| Feature | Be Well Buddy (AI Chatbot) | Be Well Texas Website (Traditional) |
|---|---|---|
| Screening Completion Rate | ✓ 32% of enrolled; 100% of those who started | ✓ 32% of those with queries; 61% completed |
| Abandonment Rate (post-start) | ✓ 3% | ✓ 39% |
| Referral Rate (of screened) | ✓ 44-83% (depending on screener) | ✓ 65% |
| Content Source | ✓ Curated closed library (expert-vetted) | ✓ Web-based informational tool |
Calculate Your Potential AI ROI
Estimate the efficiency gains and cost savings your enterprise could achieve by implementing similar AI-driven solutions.
Estimated Annual Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A phased approach to integrating AI solutions for impactful results, inspired by the pilot study's successful deployment.
Phase 1: Concept & Design
Define clear objectives for AI-driven SUD screening, focusing on user needs and clinical efficacy. Design the chatbot with a curated content library, ensuring medical accuracy, non-stigmatizing language, and compliance with healthcare regulations. Emphasize a user-centric design (e.g., PACMAD model) for optimal usability and trust.
Phase 2: Pilot Implementation & Controlled Testing
Deploy the AI chatbot in a pilot setting with a target user group (e.g., individuals at risk for SUD). Conduct rigorous testing over a defined period (e.g., 7 days) to evaluate system functionality, user engagement, and initial screening/referral metrics. Gather qualitative feedback through in-depth interviews to understand user experience and identify pain points.
Phase 3: Data Analysis & Efficacy Assessment
Analyze quantitative data on system interactions, screening completion rates, referral conversions, and technical performance (e.g., response precision, error rates). Synthesize qualitative insights to refine understanding of usability and acceptability. Assess the potential efficacy of the AI solution in achieving desired outcomes, such as increased screening and linkage to care.
Phase 4: Optimization & Scalability
Implement identified improvements based on pilot findings, addressing technical issues and enhancing user experience (e.g., direct connections to live support, localized resources). Plan for a larger-scale, randomized controlled trial to definitively establish efficacy. Develop strategies for broader deployment across diverse healthcare settings (e.g., primary care, emergency rooms) to maximize population reach and impact.
Ready to Transform Your Healthcare Services with AI?
Leverage the power of AI to enhance patient engagement, improve screening rates, and streamline access to critical services. Let's discuss how our custom AI solutions can benefit your organization.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat