2025-26 Cohort

Bio: I am a colorectal surgeon and health services researcher committed to ensuring that all patients have equitable access to high-quality colorectal cancer care. My interest grew from an observation that patients who are young, minority, or live in rural settings appeared to present with more advanced disease after having received less than the standard of care in the diagnosis and treatment of their colorectal cancer. At the same time, I could see the healthcare environment around me evolving into geographically diffuse networks of hospitals that were either merging into common healthcare systems or linked through collaborative quality initiatives. My goal is to leverage these hospital networks as a mechanism to (1) facilitate selective centralization of patients with complex oncologic needs to the highest volume centers and (2) disseminate expertise and resources to ensure most patients receive equitable high-quality cancer care closer to home. Through my roles in the Center for Health Outcomes & Policy at the University of Michigan (U-M) and the Michigan Surgical Quality Collaborative, I aim to use robust systems science methodology to create learning health systems that seek to continuously learn from and improve the quality of colorectal cancer care throughout Michigan and across the country.

Project Title: Improving Multidisciplinary Colorectal Cancer Care in a Statewide Quality Collaborative 

Project Summary: Quality measures for colorectal cancer care such as positive margins and inadequate lymph node examination vary across hospitals in the Michigan Surgical Quality Collaborative (MSQC). We are working with the MSQC to better assess quality and initiate cross-collaborative quality improvement initiatives. Because most hospitals have multidisciplinary tumor boards that help to guide clinical care for cancer patients, this may be a ideal setting to initiate quality improvement. This project seeks to build partnerships, collect baseline data, and inform this quality improvement intervention.

Bio: I am an epidemiologist focused on eye care and vision health outcomes, social risk factors, and health disparities. My career goal is to make strides towards achieving improved sensory health outcomes through research at an academic institution. My dissertation work focused on epidemiological risk factors associated with vision threatening eye disease in geographically isolated populations. My research experience related to geospatial health began during my post-doctoral training at the University of Michigan under the mentorship of Drs. Newman-Casey and Woodward. During this time, I was awarded a NIH K12 Institutional Research and Academic Career Development Award and had additional support through a NEI Diversity Supplement. I planned, coordinated, and contributed to publications on the implications of the neighborhood and built environment on eye and vision outcomes. I was able to present my work at both National and International conferences including the Association for Research in Vision and Ophthalmology, the American Public Health Association, and the American Academy of Ophthalmology annual meetings. For my research contributions, I received the Society of Clinical Trials Equity, Diversity, and Inclusion Early-Career Award and was inducted into the Edward A. Bouchet Graduate Honor Society. In 2025, I was awarded the Prevent Blindness Rising Visionary Award. 

Project Title: Implementation of a Social Risk Factor Screening for Patients with Microbial Keratitis at a Tertiary Eye Care Center 

Project Summary: Microbial keratitis (MK) is an emergent infection of the cornea causing vision impairment, acute pain, and loss of productivity. MK disproportionally affects individuals with lower socioeconomic status and those living in neighborhoods with fewer resources. These adverse social determinants of health are known as social risk factors (SRFs). MK disproportionally affects people of color and those living in poverty. Due to difficulties in obtaining care, when underserved individuals make it to the hospitals, their eye disease is often more advanced. Our group has previously reported that individuals with lower socioeconomic status come to emergency departments for MK in disproportionate numbers and, among MK patients, presenting visual acuity is significantly associated with insurance status and with living in neighborhoods with fewer resources. 

Thus, there is a critical need to screen for SRFs for MK patients and connect them with appropriate resources to improve MK outcomes. Integrating SRF screening will allow eye care providers to gain insights into the broader challenges patients may face, which can impact their ability to adhere to treatment plans, attend appointments, and achieve optimal vision outcomes. Furthermore, by leveraging social work resources, the clinic can provide targeted interventions and support services tailored to individual patient needs, thereby fostering a more patient-centered and inclusive approach to eye care delivery. 

The central hypothesis is that patients that receive SRF screening and are connected to Kellogg Eye Center/Michigan Medicine social work services will have better MK outcomes (clinic follow-up and medication access) then patients that are unable to access Kellogg Eye Center/Michigan Medicine social work services. This hypothesis is supported by our lab’s preliminary data utilizing field notes from the R01 “Quantifying Microbial Keratitis to Predict Outcomes” study that has identified accessibility (distance, no person to accompany, and transportation), affordability (medication cost, lack of insurance, and transportation cost), and accommodation (dependent care responsibilities and scheduling conflicts) SRFs that have impacted MK outcomes including follow-up and medication access that could be and have been mitigated by social work services and resources. 

The rationale for this research is to comprehensively address the holistic needs of patients beyond their ocular health concerns. By identifying and addressing patients’ SRFs, such as socioeconomic status, housing instability, food insecurity, and access to eye care, the cornea clinic may enhance patient care outcomes. This project will have the support of our multidisciplinary team including both an NIH supported cornea specialist and social worker.

Bio: As an early-career health services researcher and social epidemiologist, my work is deeply rooted in addressing health differences faced by individuals with disabilities, particularly focusing on deaf and hard-of-hearing (DHH) populations. I am also developing a research program in CHARGE syndrome, a rare genetic disorder and leading cause of congenital DeafBlindness in the United States. Motivated by a personal connection and professional commitment to the disabled community, I meet NIH's historical definition as a scientist from a disadvantaged background (NOT-OD-20-031) and am dedicated to understanding and combatting discrimination in healthcare. I am a disabled scientist, and my parents are both DHH.

My research integrates multidisciplinary approaches from epidemiology, sociology, health policy, and psychology to explore and solve issues of healthcare access, utilization, and delivery for people with disabilities. With expertise in social-behavioral aspects of public health, I am a Master Certified Health Education Specialist, which equips me to effectively develop, implement, and evaluate community-based health promotion programs. I have published over 50 peer-reviewed articles and served as Principal Investigator on multiple grants from agencies like AHRQ and the National Library of Medicine, focusing on health behaviors and information access among linguistic and cultural minority groups. My work has received recognition, with publications in Health Affairs and Patient Education & Counseling, and has influenced federal rulemaking procedures. I also contribute to the academic community through roles on the Editorial Boards of the Journal of Mixed Methods Research and the Disability and Health Journal. As an educator, I provide guidance across multiple levels of training, working with undergraduate and graduate students, dissertation committees, and post-residency fellows.

Project Title: Developing a Computable Phenotype to Identify Patients with CHARGE Syndrome 

Project Summary: CHARGE syndrome is a rare disease (1 in 10,000 births) and leading genetic cause of congenital DeafBlindness in the United States. In addition to sensory disability, CHARGE syndrome has significant impacts across organ systems. Due to incomplete penetrance (i.e., not all people with the genetic mutation develop CHARGE syndrome) and significant phenotypic variability (i.e., no clear genotype-phenotype relations), diagnosis and care coordination in CHARGE syndrome is challenging. There is no ICD-9-CM/ICD-10-CM code for CHARGE syndrome. Dr. James has developed a rule-based computable phenotype using structured ICD-9-CM, ICD-10-CM, and CPT codes related to manifestations of CHARGE syndrome. This model has incredibly poor performance, and needs to be improved before considering deployment in research and clinical settings. The objective of this study is to improve the CHARGE syndrome computable phenotype using unstructured (e.g., free-text clinical notes) data in the electronic health record, in addition to structured (i.e., diagnostic and procedure codes) data. The goal is to create a probabilistic computable phenotype using structured (ICD, CPT) and unstructured clinical data.  We anticipate that the inclusion of unstructured data will significantly improve sensitivity and specificity of the computable phenotype model, which will aid in future implementation of this algorithm in research and practice. 

• Bio: I am an ambulatory pharmacist with clinical practice experience providing care to older adults experiencing polypharmacy. I initially tried to tackle this problem by collaborating with an interprofessional team to help individual patients ‘right-size’ their medications. Through this process, I developed a strong desire to build a formal research program to study how patients make decisions about stopping or continuing their chronic medications. To further my transition towards research, I completed a part-time fellowship through the University of Michigan Center for Bioethics and Social Sciences in Medicine focused on Decision Sciences while maintaining my faculty appointment at the University of Michigan College of Pharmacy. I am currently a Multiple Principal Investigator on a NIA-funded study, “Prescribing without a guide: A national study of psychotropic and opioid polypharmacy among persons living with dementia.” 

• Project Title: Exploring decision-making preferences in older adults: Medication Engagement and Decision (MED) Screener

• Project Summary: Problem: Up to one-half of older adults take unnecessary medications (i.e., polypharmacy). Deprescribing is a strategy to address polypharmacy. However, older adults have varying attitudes towards stopping medications. Long-term goal: To develop and implement a screening tool that allows older adults to succinctly share their deprescribing preferences within the context of routine clinical visits. Primary objective: To explore older adults’ perceptions of the feasibility, usability, and relevance of the Medication Engagement and Decision (MED) Screener in clinical practice.

• Bio: I currently am a Clinical Instructor of Internal Medicine at the University of Michigan and a Research Fellow in the National Clinician Scholars Program. My research focuses on the implementation and governance of provider-facing artificial intelligence (AI) tools in the healthcare setting. During my early career and medical school training (2016-2020), my initial research focused on the development of predictive machine learning models for key clinical outcomes such as delirium, antibiotic stewardship, and perioperative outcomes. These projects engaged computational ML experts and clinical stakeholders to build and implement state-of-the-art machine learning clinical prediction models to patient care. After completing my medical degree, I began my residency in Internal Medicine (2020-2023) during which clinical AI tools became increasingly prevalent. My research transitioned toward the oversight and governance of AI models to ensure effectiveness for patients and providers. My primary work during this time was in the validation of sepsis prediction models, including a key validation study of the Epic sepsis prediction model that drew national recognition and was cited in the White House Blueprint for an AI Bill of Rights. Upon completion of my residency, I took a brief hiatus from research to serve as Chief Medical Resident (2023-2024) to build my clinical expertise. Immediately afterward, I joined the National Clinician Scholars Program (2024-2026) as a research fellow to build my research portfolio, obtain an MS in Health Care Research, and prepare for my early career development award. In my current role as a Clinical Instructor, Research Fellow, and MEL-STaR Scholar, I leverage my AI expertise to build and evaluate clinical AI technologies to improve medical diagnosis, hospital operations, and medical education.

• Project Title: Toward Explainable Sepsis Alerting Systems Using Large Language Model-Generated Clinical Reasoning

• Project Summary: Problem: Current automated sepsis alerts lack explainability, leading to high rates of provider dismissal. Limited interpretability of interruptive alerts can hinder point-of-care clinical reasoning, increase provider cognitive burden, and contribute to alert fatigue. Objective: Test the ability of large language models (LLMs) to generate clear, accurate, and context-aware clinical explanations for automated sepsis alerts to improve sepsis alert interpretability and decrease point of care cognitive burden for medical providers. Hypothesis: LLMs can generate high quality clinical explanations for sepsis alerts to contextualize real-time sepsis alerts and improve alert interpretability. A novel sepsis alerting system (SepsisEX) created using LLM-generated clinical summaries can reduce provider cognitive burden when responding to sepsis alerts in a simulated clinical setting.

• Bio: I am a Clinical Assistant Professor in Adolescent Medicine at the University of Michigan. I completed a fellowship in adolescent medicine at Stanford University and pediatrics residency at the University of Michigan.I have also obtained a master’s degree in public health, with an emphasis on health management and policy. During residency and fellowship, I gained expertise in the evaluation and management of patients with eating disorders and now these patients comprise much of my clinical practice. During fellowship my research focused on evaluating sociodemographic disparities in readmission rates for adolescent and young adult patients with eating disorders. We found that individuals with public insurance were more likely to require multiple hospital admissions. I also served as a mentor on an additional study evaluating telehealth services for patients with eating disorders. We found that quality of care was preserved for individuals who utilized telehealth services compared to in-person services.

• Project Title: Preventing eating disorders through the development of an educational toolkit

• Project Summary: Eating disorders are common and begin mostly during the adolescent and young adult years. These disorders have profound impact on both mental and physical health and have the 2nd highest mortality among psychiatric illnesses (behind opioid use disorder). As eating disorder prevalence has increased it is imperative to focus on eating disorder prevention.For this quality improvement project, we seek to develop and implement an eating disorder toolkit for caregivers that will be provided at primary care provider visits.The toolkit will be an educational resource on normal growth and development, how to talk to kids about body image, and warning signs. Additional focus will be targeted towards youth on social media engagement and when to ask for help.

Bio: I am a PhD pre-candidate in Learning Health Sciences at the University of Michigan Medical School. My research focuses on designing theoretically grounded cognitive support systems for high-stakes healthcare environments. I study how multimodal behavioral signals—including eye tracking, speech, and action dynamics—can be abstracted into interpretable cognitive structures that support clinical decision-making, team coordination, and patient safety.

My work integrates artificial intelligence and immersive technologies as implementation layers within a broader framework of real-time cognitive state modeling and evidence-based feedback. In high-risk medical contexts such as emergency resuscitation, I develop data infrastructures and analytic pipelines that transform complex behavioral streams into actionable diagnostic and debriefing tools, advancing scalable and continuously improving learning systems aligned with the principles of Learning Health Systems.

With an interdisciplinary background spanning information science, human-computer interaction, and immersive system design, I previously contributed to NSF-funded research on XR-enabled decision support and developed surgical training platforms using real patient imaging data. Building on this foundation, my doctoral research advances multimodal cognitive modeling and AI-driven feedback architectures to enhance training objectivity, reliability, and system-level learning. 

Project Title: Advancing Multi-Order Multimodal Network Modeling for Situational Awareness in XR(Extended Reality)-Based Clinical Team Simulations 

Project Summary: High-risk clinical emergencies require rapid coordination and sustained situational awareness (SA) across team members. Current evaluation approaches rely heavily on expert observation and manual coding, which are labor-intensive, difficult to scale, and limited in their ability to capture dynamic, multimodal coordination processes.

The objective of this project is to develop a methodological platform for modeling situational awareness in XR-based clinical team simulations using multi-order network representations of multimodal behavioral data (e.g., gaze, speech, interaction logs). This project builds upon Transition Network Analysis (TNA) and interpretable analytic translation (ReadGaze system). By improving the measurement and interpretability of team coordination processes, this work supports patient-centered quality improvement efforts aimed at strengthening safety infrastructures in high-risk clinical settings.

This study focuses on analytic development, feasibility validation, and interpretability assessment. It does not evaluate the comparative effectiveness of clinical interventions.

Bio: I am a doctoral student in Health Infrastructures and Learning Systems at the University of Michigan Medical School with a foundation in community health education and health equity research. My career has been dedicated to addressing health disparities among underserved and marginalized populations through community-engaged research, program development, and policy advocacy. My interest in diabetes stems from my extensive experience working with communities disproportionately affected by chronic disease disparities. Through my work at Vital Strategies and the Michigan Department of Health and Human Services, I have developed expertise in designing equity-centered strategies to reduce health disparities. I am particularly committed to leveraging community health worker models and culturally responsive interventions to improve diabetes prevention, management, and outcomes in underserved communities. My bilingual (English/Spanish) capabilities and experience conducting community-based research position me to effectively engage with diverse populations. I have demonstrated success in coordinating multi-stakeholder initiatives, developing evidence-based training curricula, and translating research findings into actionable policy recommendations. I am committed to advancing diabetes research that centers health equity and community voice. As aMcNair Research Scholar, I bring both credentialed expertise and a research-driven approach to my doctoral work. I am honored to be the recipient of the Wayne State University–Academia del Pueblo 1st Place Community Research Award, which reflects my commitment to community-centered scholarship.

Project Title: Patient Centered Stakeholder Analysis: Perspectives to BidirectionalData Exchange Between Diabetes Self-management Education(DSME) and Clinical Care

Project Summary: Despite diabetes self-management education (DSME) being a standard of medical care for all people with diabetes, lack of referral to DSME and sub-optimal patient engagement in DSME programs persists.  There is a substantial amount of evidence that demonstrates that when DSME is offered in community settings, where people live and work, outcomes, including process, clinical, behavioral, psychosocial improve. Despite this evidence,  community-based DSME programs remain largely disconnected from clinical care systems where health services are provided. This ultimately leads to information silos that impact care continuity and patient outcomes. Indeed, fewer than 5% of providers receive any feedback about patient participation, progress, or outcomes from these programs (Chomko et al., 2022). This bidirectional communication failure represents a critical gap between community-based DSME and clinical settings.

In Detroit, this gap has real implications for men who are less likely to engage in routine care and preventive services and often underutilize DSME, which can worsen diabetes outcomes. In partnership with the Detroit Health Department, an NIH-funded R01 randomized controlled trial (PI: Hawkins), is currently implementing a peer-led diabetes self-management support program for men with type 2 diabetes. Bridging community programs with clinical care is both a matter of care coordination and an access issue—ensuring that men who participate in community-based support do not “fall through the cracks” when it comes to communication with their healthcare teams. This local context underscores the need, in general, for effective data linkages between community interventions and healthcare systems.

Although health information exchange standards like Fast Health Interoperability Resources (FHIR) have advanced clinical interoperability (Mandel et al., 2016), no evidence-based framework exists for implementing bidirectional data exchange between community-based programs and clinical systems. This limits the potential scalability of community-based DSME and prevents health systems from leveraging the full potential of community-based programming that increasingly emphasize whole-person, coordinated care approaches (National Academy of Medicine, 2021).