2026 Research Grants for Penn LDI Research Initiative on Improving Primary Care for Older Adults

January 12, 2026

By:

LDI Staff

Print:

The University of Pennsylvania’s Leonard Davis Institute of Health Economics (LDI) has announced two new grant awards under its collaboration with Humana’s health services brand, CenterWell, supporting rapid-turnaround, high-impact research with measurable effects on value-based primary care and other alternative delivery models for older adults. The projects are:

See Closed RFP

Unscheduled Care Utilization for Acute Exacerbations of Chronic Disease in Value-Based Payment Models

Principal Investigator: Ari Friedman, MD, PhD | Perelman School of Medicine

Older adults are increasingly relying on emergency departments, urgent care centers, retail clinics, and telemedicine for acute health needs, yet little is known about how this shift away from primary care affects chronic disease management, overall costs, or continuity of care. Under value-based payment (VBP) models, primary care practices bear financial responsibility for total patient costs and therefore have strong incentives to manage unscheduled care effectively.

This project will leverage Humana claims data from 2017 to 2024 to investigate how non-primary care unscheduled care (NSUC) influences preventive care, chronic disease diagnosis, and downstream spending among older adults with chronic conditions. The first aim examines whether NSUC disrupts preventive care and timely diagnosis, and whether VBP arrangements buffer these effects. The second aim assesses whether NSUC increases total health care spending and whether financial risk-sharing reduces these costs, particularly through duplicate visits or higher inpatient utilization. The third aim evaluates whether primary care physician participation in VBP reduces reliance on non-primary care acute care settings, especially among patients with multiple chronic conditions.

Using longitudinal data and event-study difference-in-differences methods, this study will generate actionable evidence on how VBP design and insurance incentives can strengthen primary care, reduce fragmentation, and improve high-value acute care for older adults.

Evaluating the Safety and Appropriateness of AI-Based E-Consults for Older Adults in Primary Care

Principal Investigator: Gary Weissman, MD, MSHP | Perelman School of Medicine
Co-Principal Investigator: Matthew Press, MD, MSc | Perelman School of Medicine

Timely access to subspecialty consultation is critical for older adults in value-based care models, yet shortages and delays frequently undermine diagnostic accuracy and increase downstream costs. AI-based subspecialty e-consult systems offer the potential to improve access and timeliness, but their real-world safety, quality, and acceptability remain unknown. This proposal introduces a two-part Phase 1 evaluation to determine whether AI-generated e-consult recommendations could safely support primary care decision-making for older adults.

Aim 1 employs a discrete choice experiment involving 150 older adults and 150 primary care clinicians to quantify preferences and acceptable trade-offs among quality, cost, and turnaround time for AI-generated versus human-generated e-consults. Aim 2 conducts a randomized, blinded clinician Turing test comparing real subspecialist e-consult notes with those generated by a commercial AI tool across cardiology, nephrology, and gastroenterology. Specialists will assess quality, safety, and the ability to distinguish between human- and AI-generated outputs.

Results will clarify how good, how fast, and how reliable an AI system must be to support care in a VBP environment, while providing foundational evidence on clinical safety before any large-scale implementation. Together, these studies will inform the responsible and patient-centered deployment of AI e-consult technologies, which could improve specialist access, reduce delays, and enhance value-based primary care for older adults.