Research at PITTCARE

Department faculty and active projects — click any card to read the full detail.

Department faculty

Our researchers

OK
Olga V. Kravchenko
MS, PhD · Associate Research Director, PITTCARE
Intelligence-Tech in Primary CareIntelligence-Tech Ethics

Focuses on translating Intelligence-Technology, including machine learning, into real clinical settings — practical models, EHR-based decision support, and, most recently, ethics and fairness. Her research spans three areas: practical machine learning models translatable into family medicine and primary care; clinical decision-support tools that extract patient-specific context from EHRs; and translational work using large-scale datasets including N3C, All of Us, and the ENACT Network. Her most recent focus — the ethics of Intelligence-Technology — connects directly to the Voice of the People Survey. PhD in Chemistry, University of British Columbia; NLM postdoctoral fellowship at Pitt DBMI.

NB
Nora Bridges, PhD
VoP qualitative lead, with Rebecca Shisler & Todd Bear
Medical anthropologyVoice of the People

Leads the qualitative arm of the Voice of the People Survey, with Shisler (participatory methods) and Bear (survey methodology) rounding out the team. Dr. Bridges is an applied medical anthropologist with 15+ years of community-engaged research experience, designing the structured interviews that keep patients and clinicians active partners in shaping PITTCARE's tools. She's currently studying how ambient Intelligence-Tech tools affect patient-provider relationships across Pennsylvania practices. Rebecca Shisler (MA, PhD candidate, Sociology) brings participatory-research design to the interview arm. Todd Bear, PhD, provides the statistical and sampling architecture — including BRFSS-model survey design — that keeps VoP results representative and replicable.

RZ
Richard K. Zimmerman
MD, MPH · Vice Chair, Preventive Medicine Research
Vaccine PolicyAgent-Based Modeling

CDC ACIP voting member; with Dr. Williams, uses the FRED simulation platform to study influenza vaccination strategy — PITTCARE's most direct tie to the School of Public Health. Dr. Zimmerman chaired ACIP's influenza and adult vaccination working groups. With Dr. Katherine Williams, he applies FRED — an agent-based epidemic simulation platform from Pitt's Public Health Dynamics Laboratory — to study vaccine efficacy, dosing, and household transmission patterns. He practices at East Liberty Family Health Care Center, serving on its Quality Assurance and Ethics Committees, and is a driving force behind the department's 4 Pillars™ program.

KW
Katherine V. Williams
MD, MPH · Assistant Professor
Vaccine EpidemiologyHealth Equity

Works alongside Dr. Zimmerman on FRED-based modeling of influenza vaccine strategy by age, race, and geography. Research Project Coordinator on CDC-funded influenza studies since 2019, and APTR representative on ACIP's adult RSV vaccine workgroup since 2022. Completed a DEI fellowship in Human-Centered Design to build rural vaccine-outreach strategy — work that echoes PITTCARE's own commitment to designing Intelligence-Tech with, not just for, the communities it serves.

JM
John S. Maier
PhD, MD · Center Director, PITTCARE
Center DirectorAiM-PC curriculum

Physicist-turned-physician; directs PITTCARE and teaches the second-year critical-appraisal course on Intelligence-Technology, one of the first in the country to require the national STFM AiM-PC curriculum. PhD in Physics (UIUC) and MD; trained in Family Medicine at UPMC Shadyside, then spent nearly a decade as VP of Biomedical Research at ChemImage Corporation, co-inventing 41 patents in spectroscopic imaging. Returned to Pitt in 2011, and now directs Innovation and Strategic Initiatives for the department alongside leading PITTCARE. His translational arc — physics through industry back to academic medicine — shapes both his research and teaching.

KK
Kevin Kindler
MD · Associate Clinical Director, PITTCARE
Associate Clinical DirectorClinical Informatics

Family physician and Clinical Informatics Fellowship faculty; brings the day-to-day clinic perspective to how PITTCARE's tools actually land in a visit. Board-certified in family medicine; practices at Squirrel Hill Family Medicine, affiliated with UPMC Shadyside. Completed a Clinical Informatics Fellowship at Thomas Jefferson University after residency at UPMC Shadyside Family Medicine. Co-authored a book chapter on the ethics of AI in primary care and family medicine residency training with Dr. Visweswaran. Research interests: clinical informatics, global health, and health technology innovation.

Active projects

From research relationship to research output

Concierge Clinical Evidence Service (CCES)

DBMI · CCAI · ENACT Network
On-demand evidence synthesis for primary care clinicians, built on the national ENACT clinical data network.
CCES answers a clinician's specific patient-context question by combining structured EHR data with literature synthesis — the kind of point-of-care evidence lookup that's normally too slow to use during a visit. It draws on ENACT's federated network of clinical data across participating health systems, with NLP infrastructure contributed by Pitt's PittNAIL lab.

Ambient Intelligence-Tech & the Patient-Provider Relationship

Qualitative study · Dr. Bridges
A qualitative study of how ambient documentation tools (often called ambient AI) change clinical workflow and the patient-provider relationship across Pennsylvania family medicine practices.

Multi-Agent LLM for Secondary Headache Diagnosis

Dr. Wang, with DBMI & UT Southwestern
A multi-agent LLM system — one orchestrator, seven specialist agents — for flagging secondary headache in primary care, tested on 90 expert-validated cases (F1 = 0.605).
Agents cover thunderclap onset, meningismus, papilledema, temporal arteritis, systemic illness, focal neurologic deficits, and first-or-worst headache presentation. The MERLIN prototype is publicly available as a Docker image.

Agent-Based Modeling of Influenza Vaccination (FRED)

Zimmerman & Williams · School of Public Health
Simulating population-level flu burden under different vaccination strategies, using Pitt's FRED agent-based epidemic platform.
FRED represents every person in a region individually, inside real household, school, and workplace networks — letting Zimmerman and Williams model questions no clinical trial can answer directly, like the effect of uneven household vaccination or two-dose regimens for older adults. 10+ FRED-based publications since 2010.
2010–Ongoing collaboration
10+Publications

4 Pillars™ Practice Transformation Program

Dept. of Family Medicine
A home-grown, evidence-based model for closing the gap between what clinicians know and what happens in a visit — organized around access, communication, systems, and motivation.
Developed to protect patients from vaccine-preventable disease, the program was tested in before-and-after studies, then randomized cluster trials, then put into clinical use across UPMC — and found cost-effective. 14 peer-reviewed publications; National Publication Award from the National Adult and Influenza Immunization Summit. For PITTCARE, it's a template for how quality-improvement infrastructure already trusted by clinicians can carry new intelligence-technology into practice.
14Publications
4Evidence-based domains

Critical Appraisal in the Setting of AI — Course

Pitt Med Year 2 · Dr. Maier
A required second-year course teaching evidence-based critical appraisal applied specifically to Intelligence-Technology — one of the first in the country to require the national STFM AiM-PC curriculum.
All enrolled students complete the first two modules of the STFM Artificial Intelligence and Machine Learning for Primary Care (AiM-PC) curriculum. This satisfies the Education and Workforce domain of the STFM Family Medicine AI Centers of Excellence recognition program, and gives every Pitt med student an early, structured framework for evaluating Intelligence-Technology in clinical practice — not just using it.

FM PittNet

CTSI · Practice-Based Research Network · Est. 2023
A practice-based research network connecting primary care patients, practices, and investigators — PITTCARE's real-world pipeline for testing Intelligence-Tech research in actual clinics, not just simulated settings.