A Comparison Between Online Communities and AI: Effects on Healthcare Decisions and Outcomes

Presentation Time: 09:15 AM - 09:30 AM

Abstract Keywords: Artificial Intelligence/Machine Learning, Social Determinants of Health, Patient-Generated Data / Patient Reported Outcomes (PROs)
Primary Track: AI and Care Outcomes
Programmatic Theme: Emerging Technology and Technical Infrastructure

Orthopedic treatment effects are heterogenous in the U.S., contributing to a sense of uncertainty for patients about the best treatment option. This study compares personalized responses from online communities with AI responses to assess how both address hedonic and utilitarian motivations in orthopedic treatment decisions. We explore the role of Large Language Models to respond to clinical questions based on patient-driven motivations to support orthopedic treatment shared decision making.
Speaker(s):
Benjamin Schooley, PhD
BYU

Author(s):
Benjamin Schooley, PhD - Brigham Young University;

Interpretable Machine Learning to Identify Risk Factors for Recidivism in Intimate Partner Violence

2025 Clinical Informatics Conference DEI/Health Equity Presentation

Presentation Time: 09:30 AM - 09:45 AM

Abstract Keywords: Artificial Intelligence/Machine Learning, Algorithmic bias and impacts on Health Equity, Diversity, Equity and Inclusion, Data Science
Primary Track: AI and Care Outcomes
Programmatic Theme: Informatics-Driven Value-Based Healthcare

Predicting recidivism in Intimate Partner Violence cases remains a critical challenge in prevention. We present a filtered target encoding technique that standardizes clinician-assigned severity scores and captures non-linear relationships between risk factors and recidivism. Applied to data from a four-year clinical study, our approach reveals decreased substance use and marital status significantly influence reoffending risk. By combining perpetrator and partner-reported variables, we enhance risk assessment accuracy and enable effective interventions for preventing IPV recurrence.
Speaker(s):
Cerag Oguztuzun, PhD
Case Western Reserve University

Author(s):
Cerag Oguztuzun, PhD - Case Western Reserve University; Mehmet Koyuturk, PhD - Case Western Reserve University; Gunnur Karakurt, Psychiatry - Case Western Reserve University;

Guided Care: Personalized, AI-Driven Care Journeys to Enhance Patient Engagement and Outcomes in Value-Based Care Models

Presentation Time: 09:45 AM - 10:00 AM

Abstract Keywords: Care Delivery Models, Artificial Intelligence/Machine Learning, Adaptive Clinical Decision Support
Primary Track: AI and Care Outcomes
Programmatic Theme: Informatics-Driven Value-Based Healthcare

Guided care offers a patient-centered, structured approach that transforms clinical guidelines into personalized care journeys. This model, tailored for various conditions, utilizes AI to stratify patients by health status and risk, aligning care activities with individual needs. Patients enrolled in guided care journeys benefit from integrated digital and in-person touchpoints, managed by a dedicated care team that includes healthcare providers, nutritionists, and wellness experts. Using AI-powered segmentation and continuous monitoring, the care team designs interventions around specific patient goals and outcome metrics, such as PROMs and PREMs, ensuring dynamic adaptability to patient progress. By tracking these indicators, the platform enables value-based care delivery, emphasizing improved engagement, proactive care, and long-term outcomes. Guided care exemplifies a scalable, structured approach to personalized, goal-driven care that supports healthcare providers in delivering comprehensive, value-based services to diverse patient populations.
Speaker(s):
Mohammad Ghosheh, M.D
iO Health

Author(s):

From Prediction to Practice: Learnings from a Successful AI-based Intervention to Improve End-of-Life Care

Presentation Time: 10:00 AM - 10:15 AM

Abstract Keywords: Artificial Intelligence/Machine Learning, Adaptive Clinical Decision Support, Communication Strategies, Risk Measurement, Communication Strategies, Health IT Standards (USCDI, FHIR®, SMART, etc.), Workflow Efficiency, Clinical Process Automation
Working Group: Clinical Decision Support Working Group
Primary Track: AI and Care Outcomes
Programmatic Theme: Clinical Decision Support and Analytics

Generating accurate AI-based predictions is one thing, but getting clinicians to act on these predictions is another. We describe the key strategies to engage and empower clinicians from our clinical decision support system that used machine learning to predict mortality risk and improve end-of-life care. Using clinical training, streamlined workflows and engaging messaging, we observed markedly high response rate to prediction alerts (94%) and system-wide improvements in palliative care metrics over 4 years of implementation.
Speaker(s):
Nathan Moore, MD
BJC

Author(s):
Jessica Saleska, PhD - Washington University in St Louis / Central Health Intelligence; Nathan Moore, MD - BJC;