Citations

Objectives/Goals: This project will enrich our understanding of basal breast carcinogenesis, highlighting the distinct biological differences in breast cancer risk between non-Hispanic Black (NHB) or non-Hispanic White (NHW) women, but also has the potential to provide real-world solutions thereby contributing to reducing health disparities in breast cancer outcomes Methods/Study Population: In the biological investigation, I am utilizing a unique cohort of normal breast tissues from Mayo Clinic patients who underwent breast reduction surgery that have self-identified as NHB or NHW premenopausal women. From these tissues, we are able to subculture human mammary epithelial cells, which we will use for our consequent experiments. For the qualitative study, I aim to recruit up to 40–50 eligible NHB women identified at increased risk for breast cancer and conduct semi-structured qualitative interviews informed by the National Institute on Minority Health and Health Disparities Research Framework. Results/Anticipated Results: Gene expression profiling on our cohort of 15 NHB and 40 NHW premenopausal women identified a gene expression signature in NHB women indicative of elevated Hedgehog signaling, a key factor in triple-negative breast cancer (TNBC) development and progression. I hypothesize that persistent activation of Hedgehog signaling within NHB women instigates the proliferation and transformation of basal stem cells within the breast, thus fueling TNBC development. Understanding the perceptions and experiences of NHB women that are identified as being at an increased risk for BC will aid in the identification of barriers and facilitators to their risk reduction care and in turn provide implementation strategies that could contribute to alleviating the racial disparity seen in TNBC morbidity, mortality, and preventative care. Discussion/Significance of Impact: The significance of this project lies in its dual approach, uncovering basal breast cancer mechanisms in a high-risk group using normal breast tissue and evaluating the perspectives from NHB women to enhance the practical relevance of our findings.

Objectives/Goals: Rural disparities in chronic disease burden are well documented. However, the impact of individual socioeconomic status on these rural disparities in prevalence and control status of chronic diseases remains less understood. We aim to assess the association of prevalence and control status of chronic diseases among adults with rurality and socioeconomic status (SES). Methods/Study Population: The Expanded Rochester Epidemiology Project medical records linkage system identified prevalence of asthma, diabetes, hypertension, and mood disorder using ICD codes between 2014 and 2019 among adults in 27 Upper Midwest counties. Uncontrolled status was defined by presence of ED visit or hospitalization with corresponding ICD-9/10 codes. SES was measured by HOUSES index, a validated individual-level SES measure, and rural status was defined by primary Rural Urban Commuting Area codes of 4–10. Hierarchical logistic regression models were used to examine the association of rurality with prevalence and control status of the four chronic diseases, adjusting for age, sex, race/ethnicity, and presence of general medical exam extracted by CPT codes (Model 1) and additionally SES measured by HOUSES (Model 2). Results/Anticipated Results: Among 455,802 adults, 42.8% were rural residents and 87.4% were Non-Hispanic White. Model 1 (without HOUSES) showed higher prevalence and lower control rates of chronic diseases in rural residents, except for asthma. In Model 2, accounting for SES, urban residents were more likely to have an uncontrolled mood disorder, hypertension, and diabetes, and there were no differences in prevalence of chronic diseases that existed between rural and urban residents. Lower SES measured by HOUSES was consistently and significantly associated with higher prevalence and uncontrolled status of chronic diseases (p-value Discussion/Significance of Impact: Patient’s SES explains rural disparities in prevalence (null association with rurality once SES is accounted) and changes the directionality of association for uncontrolled status in Upper Midwest, highlighting the importance of considering SES in rural disparities research.

Objectives/Goals: Our study’s objective is to evaluate RadOnc-GPT, a GPT-4o powered LLM, in generating responses to in-basket messages related to prostate cancer treatment in the Radiation Oncology department. By integrating it with electronic health record (EHR) systems, the goal is to assess its impact on clinician workload, response quality, and efficiency in healthcare communication. Methods/Study Population: RadOnc-GPT was integrated with patient EHRs from both hospital-wide and radiation-oncology-specific databases. The study examined 158 pre-recorded in-basket message interactions from 90 non-metastatic prostate cancer patients. Quantitative natural language processing analysis and two randomized single-blinded grading studies, involving four clinicians and four nurses, were conducted to evaluate RadOnc-GPT’s response quality in completeness, correctness, clarity, empathy, and estimated editing time. Response times were measured to estimate the time saved for clinicians and nurses. The study population included patient messages across all phases of care (pre-, during, and post-treatment) for those undergoing radiotherapy. Results/Anticipated Results: In the single-blinded grader study, clinician graders evaluated 316 responses (158 from human care teams and 158 from RadOnc-GPT). Results showed RadOnc-GPT outperformed human responses in empathy and clarity, while humans excelled in completeness and correctness. Sentiment analyses using TextBlob and VADER revealed RadOnc-GPT responses had a positive mean score of 0.25, whereas human responses clustered around neutral. VADER analysis indicated a high median score for RadOnc-GPT, nearing 1.0, reflecting predominantly positive sentiment, while human responses displayed a broader sentiment range, indicating sensitivity to context. Clinicians averaged 3.60 minutes (SD 1.44) to respond, compared to 6.39 minutes (SD 4.05) for nurses, highlighting RadOnc-GPT’s efficiency in generating timely responses. Discussion/Significance of Impact: RadOnc-GPT effectively generated responses to individualized patient in-basket messages, comparable to those from radiation oncologists and nurses. While human oversight is still necessary to avoid errors, RadOnc-GPT can speed up response times and reduce pressure on care teams, shifting their role from drafting to reviewing responses.

Objectives/Goals: Undergraduate Medical Education (UME) may apply Just-in-Time training (JITT) to provide medical students with learning experiences closely aligned with real-time clinical needs. The purpose of this scoping review is to offer an overview of the implementation of JITT training in UME. Methods/Study Population: Following the five-stage framework by Arksey and O’Malley to methodically collect and analyze studies on JITT in UME, five electronic databases were searched, and a supplemental search for grey literature was conducted. Studies exploring the integration of JITT principles into UME clinical training and their time to follow-up after training were included. Bloom’s Taxonomy was used to assess educational goals of JITT interventions. Results/Anticipated Results: The review yielded 21 studies across 4 countries. The majority were cohort studies (13) and randomized control trials (5). Assessment definitions and use of JITT varied widely. Most studies focused on short-term outcomes, defined by being measured immediately after JITT session (15) or at the end of JITT-based rotation or clerkship (3). Three studies evaluated outcomes at a period longer than 2 weeks after completion of session or clerkship. Attitudes (9), followed by skills (8) were the most common educational goals of intervention. The efficacy and utility of JITT in improving educational goal acquisition was demonstrated in 90% (17/19) of the studies with reported outcomes. Discussion/Significance of Impact: The introduction of JITT in UME has been shown to meet the immediate needs of healthcare environments; however, evidence is limited in the evaluation of longer-term outcomes. Further research to determine the impact of JITT on long-term learning retention and education goal acquisition in UME is merited.

Objectives/Goals: To assess the attitudes toward social justice of residents and faculty following the didactic series and evaluate the impact of the longitudinal curriculum on resident engagement in social justice activities through community engagements. Methods/Study Population: Data were gathered through questionnaires collected from participating residents and faculty. The session topics were chosen based on “Precepting Toward Social Justice” curriculum from the University of Minnesota Department of Family Medicine and Community Health, including sessions on advocacy, diversity, and implicit bias, among others. The self-assessment questionnaires were specific to the topics of the lecture, assessing knowledge, skills, and attitudes before and after the education sessions. Participants were asked to rate these aptitudes of each topic from “none” to “expert.” Results/Anticipated Results: Out of 22 total participants, 16 returned surveys. Overall, there was a substantial improvement in aptitudes after the lecture. There was an increase in reported above average knowledge of topics: Agency (12.5% to 56.25%), Humility (18.75% to 62.5%), Dignity (12.5% to 62.5%), Partner (18.75% to 50%), Awaken (37.5% to 75%), Place (12.5% to 43.75%), Asset (0% to 37.5%), Sanctuary (6.25% to 50%), Belonging (0% to 37.5%), Whole (0% to 43.75%), Liberation (12.5% to 68.75%), and Heal (37.5% to 75%). 31% reported that the lectures alone had an above average influence on pursuing more social justice activities through community engagement. 44% reported that the lectures had an above average impact on their subsequent patient encounters. 69% stated that this series helped develop their overall social justice skills. Discussion/Significance of Impact: All patients deserve equitable care. Family physicians have a unique role in addressing social justice concerns to eliminate inequalities and drive for better health outcomes. By enforcing social justice in education and implementing it into practice, satisfaction, trust, and well-being of patient and provider will be improved.

Objectives/Goals: This proposal outlines the successful deployment of a research training initiative to support the formation of a Learning Healthcare System. Mayo Clinic Health System (MCHS) rural providers were offered the opportunity to the fundamentals of clinical research via Clinical and Translational Science Awards core curriculum, mentorship, and an online seminar series. Methods/Study Population: MCHS funded 4 key introductory research courses: 1) Manuscript Writing, 2) Grant Writing, 3) Basic Biostatistics, and 4) Essentials of Clinical and Translations Science Program. In addition to course offerings, a Research Interest Group was formed to guide novice rural researchers on topic selection and study design. This cultivated interest to create a 16-month clinical research webinar series offering CME credits. Subsequently, an internal MCHS RFA was launched seeking early-stage investigator pilot proposals focused on rural health. Results/Anticipated Results: In 2023, over 140 MCHS providers enrolled in 324 CCaTS research courses. This training led to the submission of 53 proposals to the inaugural MCHS 2023 RFA, of which 15 were awarded. Additionally, 14 MCHS extramural grants were submitted in 2023. Training efforts expanded in 2024 to include an online research seminar series covering various study topics and providing CME credit, with an approximate attendance up to 196 attendees per session. The second annual MCHS RFA resulted in 4 internal awards, with an additional 22 extramural grant submissions. These collective efforts have increased the number of MCHS first and last author publications and the number of MCHS providers with academic rank. Discussion/Significance of Impact: Leadership’s commitment of resources to educate, mentor, and engage clinicians was crucial to our success and demonstrated a strong return on investment. To maximize impact in community-based practice, continued commitment is needed in the form of protected research time, funding, and research administration support of projects of interest

Objectives/Goals: The expanding emphasis on translational science necessitates a rethinking of traditional academic formats. To align with the central themes of CTS, we have redesigned our PhD journal club and WIP sessions, introducing novel and innovative approaches that enhance the relevance of these activities to real-world scientific and clinical challenges. Methods/Study Population: The newly adapted journal club format for CTS Predoctoral students at Mayo Clinic maintains the traditional focus on literature review but now incorporates a structured analysis of the clinical implications and potential applications of the research. This innovation aims to foster a deeper understanding of how basic research findings can be translated into improved patient outcomes and healthcare practices. Similarly, the WIP sessions have been restructured to offer an engaging and dynamic learning environment designed to empower clinical and translational science predoctoral students to effectively present their research while emphasizing the challenges they have overcome, demonstrating the translational potential of their findings, and enhancing their communication skills. Results/Anticipated Results: Feedback from participants demonstrates strong support for the new format. Students report a greater engagement with the material and a clearer understanding of how their research can contribute to improving patient outcomes. Discussion/Significance of Impact: These changes accommodate the diverse projects in CTS and embody a commitment to pushing the boundaries of knowledge in CTS. This dual transition marks a significant advancement in preparing PhD students for careers in translational science, ensuring that their research is not only rigorous but also impactful in the real world.

Objectives/Goals: Community-based practices have limited research opportunities for providers. As a rural community-based health system, Mayo Clinic Health System (MCHS) sought to create intramural funding mechanisms to help early-stage researchers conduct pilot research so they are more competitive to compete for more robust extramural funding opportunities. Methods/Study Population: We created a Research Seed Grant Program infrastructure across our four MCHS regions (Southwest and Southeast Minnesota, Northwest and Southwest Wisconsin). This model program featured an initial research funding announcement call based on 6 prioritized programs (AI Validation & Stewardship, Cancer, Health Equity, Population Health, Rural Health, Learning Healthcare System [LHS]), with submissions uploaded to a central electronic repository. Proposal review and ranking was organized on a regional basis, with ranking of all proposals by at least 2 clinical scientist reviewers according to the NIH domain-specific framework (i.e., 1.0 best > 9.0 worst). Awards were competitively selected by conformity to prioritized research areas and through ranking of most competitive overall application scores. Results/Anticipated Results: For our inaugural RFA, we received 55 grant application submissions across the MCHS regions. Fifteen of the most highly ranked applications were selected for awards on a per region basis, providing direct funding as well as protected investigator research time of 10% for up to six months. The selected projects address several research priorities including improving access, reducing health disparities, improving behavioral health in our communities, increasing cancer screening and prevention, and community-based pragmatic trials and interventions. Outcomes from these now completed pilot projects remain pending at this time. Funding for this seed grant program was supported by philanthropy, Mayo Clinic Research Administration, Mayo Clinic Comprehensive Cancer Center, and the Mayo Clinic CCaTS Rural Core. Discussion/Significance of Impact: We present this framework for a LHS-focused Seed Grant program model for consideration of adoption by other national/international LHS. Future plans include tracking of outcome metrics (e.g., published peer review articles, extramural grant applications) of this initial cycle and future expansion of this program to support the goals of our LHS.

Objectives/Goals: Longitudinal studies capture dynamic host–environment interactions crucial to microbiome research. Commonly used mixed-effects models can struggle with small clinical samples, like many endometrial cancer studies. We will compare them to simpler fixed-effects models, aiming to develop a new fixed-effects differential abundance alternative. Methods/Study Population: Starting with simulated data for linear models with a single treatment effect, we will incrementally increase complexity to reflect real-world microbiome data. To observe conditions of misspecification, simulated data of repeated measures will be generated using different distributions, while varying parameters and inter-class correlation. In each condition, linear fixed- and mixed-effects models will be fitted 1000 times to empirically determine p-value, type 1 error rate, and power. We will increase the verisimilitude of our approach by using endometrial cancer data from our lab as a real background, while adding simulated signals. This data will be used to assess extant differential abundance models against our novel fixed-effects model. Results/Anticipated Results: Our initial findings indicate that fixed-effects models maintain control over type 1 error rates in small samples, while mixed-effects models do not perform as reliably. We are now exploring the effects of confounders on type 1 error rates and power in linear models, while our next step will evaluate generalized linear models in a differential abundance context. We expect that the fixed-effects models will continue to be as reliable as mixed-effects models, while being less computationally complex. After establishing our theoretical basis, our novel model should perform comparably to mixed-effects models on pseudo-simulated endometrial cancer data and better at small sample sizes. We will finally apply our model to real endometrial cancer data to identify microbial markers of disease predisposition. Discussion/Significance of Impact: By characterizing and improving these statistical tools, we seek to increase the robustness and power of microbiome study results (especially at small sample sizes), ultimately supporting better-informed translational research decisions. This work underscores the importance of statistical methods in advancing microbiome research.

Objectives/Goals: Manual skin assessment in chronic graft-versus-host disease (cGVHD) can be time consuming and inconsistent (>20% affected area) even for experts. Building on previous work we explore methods to use unmarked photos to train artificial intelligence (AI) models, aiming to improve performance by expanding and diversifying the training data without additional burden on experts. Methods/Study Population: Common to many medical imaging projects, we have a small number of expert-marked patient photos (N = 36, n = 360), and many unmarked photos (N = 337, n = 25,842). Dark skin (Fitzpatrick type 4+) is underrepresented in both sets; 11% of patients in the marked set and 9% in the unmarked set. In addition, a set of 20 expert-marked photos from 20 patients were withheld from training to assess model performance, with 20% dark skin type. Our gold standard markings were manual contours around affected skin by a trained expert. Three AI training methods were tested. Our established baseline uses only the small number of marked photos (supervised method). The semi-supervised method uses a mix of marked and unmarked photos with human feedback. The self-supervised method uses only unmarked photos without any human feedback. Results/Anticipated Results: We evaluated performance by comparing predicted skin areas with expert markings. The error was given by the absolute difference between the percentage areas marked by the AI model and expert, where lower is better. Across all test patients, the median error was 19% (interquartile range 6 – 34) for the supervised method and 10% (5 – 23) for the semi-supervised method, which incorporated unmarked photos from 83 patients. On dark skin types, the median error was 36% (18 – 62) for supervised and 28% (14 – 52) for semi-supervised, compared to a median error on light skin of 18% (5 – 26) for supervised and 7% (4 – 17) for semi-supervised. Self-supervised, using all 337 unmarked patients, is expected to further improve performance and consistency due to increased data diversity. Full results will be presented at the meeting. Discussion/Significance of Impact: By automating skin assessment for cGVHD, AI could improve accuracy and consistency compared to manual methods. If translated to clinical use, this would ease clinical burden and scale to large patient cohorts. Future work will focus on ensuring equitable performance across all skin types, providing fair and accurate assessments for every patient.