In This Story
When physicians want to leverage complex or large healthcare data, deciding where to begin or where to dive deeply can be challenging. Obtaining quality data, whether by designing new studies, recruiting patients, or extracting data from EHRs or national databases, and turning raw data into meaningful insights requires time, specialized expertise, a, and the right tools and resources, which are limited for busy clinicians, even with the rise of ChatGPT. Statisticians at George Mason University’s College of Engineering and Computing have stepped up to help.
Through the institutional partnership between GMU Statistics Department and Inova Health System, established in 2019, faculty statisticians collaborate with physicians to address real-world clinical or health challenges. This partnership was supported by an Inova contract funded through a National Institutes of Health (NIH) National Center for Advancing Translational Sciences award from 2019 to 2024, and has been fully supported by Inova Research Administration since February 2024. The collaboration brings together evolving groups of statisticians and data scientists with physicians across multiple Inova institutes, advancing both medical and statistical/AI research. It supports both junior and established researchers in translating complex medical data into insights that improve patient care.
At the center of today’s GMU-INOVA collaboration core is a faculty team led by PI Jiayang Sun with co-PIs Brett Hunter, Ben Seiyon Lee, Nicholas Rios, Tokunbo Fadahunsi, and Isuru Dassanayake. These statisticians collaborate or provide research support on projects ranging from study design to advanced modeling, proposal, and publication. Physicians submit research requests through a structured intake process, and projects are reviewed and matched with faculty based on expertise, with some student help – which also provides valuable training to our students.
The result is a collaboration that blends productivity with rigor, ensuring research moves forward without sacrificing quality.
Assessing risk, refining reference ranges, and improving education
Rios is partnering with an Inova physician to develop a statistical model for screening patients for aortic stenosis, a serious heart condition caused by narrowing of the aortic valve. Using patient data, the project seeks to identify a small set of indicators such as age, blood pressure, and heart measurements that can predict disease risk. He has evaluated multiple modeling approaches to produce a tool that is both reliable and clinically interpretable.
Fadahunsi is currently working on a pediatric study to establish reference ranges for newer cardiac biomarkers, high-sensitivity troponin I and NT-proBNP, which are increasingly used in clinical practice. In another project with Inova’s OB/GYN department, he is evaluating how a structured curriculum can improve the use of Ambient AI Scribe technology in residency training.
Dassanayake has contributed to several projects focused on patient outcomes and clinical training. His work includes a retrospective study on malnutrition among hospitalized pediatric patients to identify patterns that can support earlier intervention, as well as a longitudinal analysis of pediatric emergency medical services in Fairfax County to better understand utilization and outcomes. He has also supported research on medical education, including simulation-based training to improve communication with patients who have limited English proficiency and studies examining how adverse childhood experiences may influence the timing of autism diagnoses.
Hunter has worked with orthopedic surgeons on two projects: one to a study on reoperation rates classified by risk factors and screw types used in femoral neck fracture repairs, and one examining the costs associated with different surgical procedures. He has also worked on medical education studies with Inova’s OB/GYN department, including a study examining the wellness of their residents, a study on incorporating palliative care curriculum into residency education, and a study on the comfort level of their residents in assisting with family planning related care.
Lee has collaborated on a range of retrospective clinical studies with cardiologists, emergency medicine physicians, nurses, intensivists, and extracorporeal life support (ECLS/ECMO) specialists, especially through collaborations with the Inova Schar Heart and Vascular Institute. He has provided statistical analyses and guidance on multiple studies, including racial differences in allograft injury among heart transplant recipients, outcomes among patients with cardiogenic shock, differences in outcomes between hub and spoke hospitals within regional cardiogenic shock networks, and the impact of early management bundles for severe sepsis and septic shock on patient outcomes. He has also involved graduate students in these collaborations, providing opportunities to apply statistical methods to real-world healthcare problems.
Sun has collaborated on a bootcamp for physicians, written letters of support, led/collaborated on grant proposals, and various other projects. They included a clinical trial revision submitted for FDA approval; a cross-sectional investigation of SARS-CoV-2 seroprevalence and associated risk factors in children and adolescents in the US; the Effect of Obesity and Obstructive Sleep Apnea on Cardiac and Inflammatory Biomarker Profile of Patients with Atrial Fibrillation; Cardiovascular Proteomics Profiles in Replacement and Interstitial Myocardial Fibrosis: The Multi-ethnic Study of Atherosclerosis; Improving Pediatric Readiness, Development of Decision Support for Personalized Treatment Recommendation in Older Adults with Coronary Artery Disease Using NLP; and the development of interpretable ML models to support heart transplant Donor-recipient matches.
A symbiotic partnership
Across all projects, statisticians play a critical role not only in analyzing data but also in shaping how findings are interpreted and communicated. Part of the work involves collaborating with physicians cut through complex datasets and focus on what matters most for patient care. This translation of analysis into practical insight is essential for clinical decision-making.
The partnership also benefits faculty professionally. Participants are compensated for their work and gain opportunities to co-author publications and contribute to grant-funded research. For example, Fadahunsi and Lee have co-authored studies examining cardiogenic shock outcomes, metabolic factors linked to cognitive performance, and the effectiveness of sepsis care protocols. Dassanayake co-authored a study comparing two approaches to coronary angiography. Collectively, this work demonstrates the collaboration’s reach across specialties and its ability to inform both clinical practice and health systems research. This collaborative core also provides statistics graduate students with opportunities to engage in interdisciplinary research with real-world clinical impact.
For many involved, the most meaningful outcome is the real-world impact, applying statistical expertise in ways that directly improve patient health and clinical decision-making.
In this rapidly changing environment, expertise, sound judgment, modern technical and communication skills, and statistical critical thinking are more important than ever.