News and Events
Chair’s Message
We are moving toward our vision with a number of activities across our various programs. We have updated our strategic plan in response to the 10-year academic program review that we recently completed. For our research-oriented MS and PhD programs, we have recently added a specialization in Data Science. We are completing a curriculum revision for our on line applied clinical informatics MS which will be effective Fall 2020. The work of our fellows in the clinical informatics fellowship program has received plaudits from clinical administrators and faculty, and we are currently recruiting a new faculty member in our department to assist with this program (view position description). We are also recruiting a faculty member in medical education to start Summer 2020 (view position description). This is the beginning of a new cycle of admissions to our graduate programs, and we look forward to another productive year, and new growth in our department.
Cordially,
Peter Tarczy-Hornoch, MD
Chair and Professor, Department of Biomedical Informatics and Medical Education
Biomedical Informatics and Medical Education Newsletter
April 22 – April 26, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Casey Overby Taylor, Ph.D. (she/her)
Thursday, May 2nd – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: Detecting and Mitigating Telehealth-Generated Inequities
Abstract: Although telehealth holds great promise to provide better access to care and improved clinical outcomes, for some groups, the benefits may not be fully realized. Here we will discuss work studying differences in patterns of telehealth adoption in terms of health equity determinants. We will also describe lessons learned from our work piloting an intervention to improve access to video visits for underserved patient groups.
Speaker Bio: Casey Overby Taylor is Associate Professor of Medicine – General Internal Medicine (GIM) and Biomedical Engineering in the Johns Hopkins (JH) School of Medicine, Associate Director of the JH Institute for Computational Medicine, and member of the JH Malone Center for Engineering in Healthcare. She is affiliated with the GIM Biomedical Informatics Data Science Section, and has joint appointments in the Dept. of Health Policy and Management in the Johns Hopkins Bloomberg School of Public Health, and the Computer Science Dept. in the Johns Hopkins Whiting School of Engineering. Her research draws from biomedical informatics and the related field of biomedical data science, to address the challenge of how to incorporate digital health technologies into clinical practice. The mission of her research group, Translational Informatics Research and Innovation (TIRI) Lab, is to understand and create advanced technology and digital device solutions that address challenges to the translation of biomedical data science-informed guidance into clinical use to improve the health of individuals, especially for people that are often underrepresented in research.
UPCOMING EXAMS
Title: Assessing Disparities Through Missing Race and Ethnicity Data: Results from a Juvenile Arthritis Registry
Student: Katelyn Banschbach
Date/Time: Monday 4/29 at 2:30-3:50
Location: Zoom only – https://washington.zoom.us/my/peter.th
Abstract: Racial and ethnic minorities remain underrepresented in research despite similar willingness to participate. Incomplete race and ethnicity data can lead to exclusion in analysis and those missing this data are more likely to be Black or Hispanic, further worsening disparities. Research and secondary analytics done with incomplete race and ethnicity can unintentionally worsen disparities. Alternatively, missing data may obscure disparities which are already present. Ensuring high quality race and ethnicity data within the EHR and across linked systems, such as patient registries, allows identification of disparities and is necessary to achieve a goal of inclusion of racial and ethnic minorities in scientific research.
Missing race and ethnicity data was assessed and completed within Pediatric Rheumatology Care Outcomes Improvement Network (PR-COIN). The project consists of 4 Aims: (1) Identifying baseline missing race and ethnicity data, (2) Understand current race and ethnicity collection practices and entry into the registry at each center via a REDCap survey, (3) Data completion through three audit and feedback cycles where reports of patients with missing data are sent to each center with request for manual completion via EHR data, (4) Impact assessment on outcome measures via comparison of racial and ethnic differences in risk of certain outcome measures such as elevated clinical juvenile arthritis disease activity score (cJADAS) which are compared pre and post data completion.
The PR-COIN database contains over 5,000 active patients with juvenile idiopathic arthritis spanning 50,000 encounters with plans to add more pediatric rheumatologic diseases over time. Completing missing race and ethnicity data will help avoid unintentionally building inequitable algorithms and system frameworks. We describe the process of identifying and completing missing race and ethnicity data at six centers within the PR-COIN network and highlight the impact of completed data on outcome assessments.
April 15 – April 19, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Brody H Foy, DPhil
Thursday, April 25th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: Mathematical and computational frameworks for adaptively benchmarking patients
Abstract: When a patient has bloodwork analyzed, results are typically benchmarked against crude, population-level definitions of normality, without consideration of what ‘healthy’ means for the given patient, in their given context. In this talk, I will overview various computational approaches we have developed to construct adaptive, and patient-specific definitions of ‘normality’. I will illustrate how these approaches can help improve diagnostic and prognostic evaluation, while also leading to a variety of novel mechanistic insights into human physiology.
Speaker Bio: Dr. Foy is a mathematician whose research is focused on improving usage of routine clinical data sources. His lab develops computational tools for high-throughput analysis of laboratory test results and associated raw data streams, with particular emphasis on hematology, and aims to build deployable tools, which are useful in both high- and low-resource settings. He did his DPhil in computer science at the University of Oxford, as a Rhodes Scholar, and undertook postdoctoral training in systems biology at Harvard Medical School.
PAPERS & PRESENTATIONS
Yeon Mi Hwang, Qi Wei, Samantha Piekos, Bhargav Vemuri,Sevda Molani, Philip Mease, Leroy Hood, Jennifer Hadlock. Maternal-fetal outcomes in patients with immune-mediated inflammatory diseases, with consideration of comorbidities: a retrospective cohort study in a large U.S. healthcare system. Lancet eClinicalMedicine. Feb 1, 2024. https://doi.org/10.1016/j.eclinm.2024.102435
ANNOUNCEMENTS
Oliver Bear Don’t Walk IV has received a K99/R00 grant from NLM titled “Collaboratively Identifying Population-Specific Social Determinants of Health for Indigenous Patients Living with HIV: From Patient Perspectives to the Electronic Health Record.” Andrea Hartzler, Meliha Yetisgen, Heidi Crane, Vanessa Simonds, Jason Deen, and Peter Tarczy-Hornoch will provide mentorship. The project aims to will connect Indigenous knowledge on social determinants of health (SDH) with the EHR.
April 8 – April 12, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Taryn Hall, MPH PhD
Thursday, April 18th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: Understanding payer coverage requirements for novel health technologies
Abstract: Reimbursement is a crucial step in translating medical advances from bench to bedside. In this talk we will discuss how payers view novel health technologies and the evidence required for payer coverage. Additionally, we will cover common missteps innovators make as they attempt to bring new discoveries to market.
Speaker Bio: Dr. Taryn Hall is a Senior Director at Optum Genomics, a precision medicine team within Unitedhealth Group (UHG) — a large healthcare company that includes a payer and healthcare services. As a subject matter expert, Dr. Hall provides precision medicine strategy consulting and innovative health program and product design for stakeholders throughout the UHG enterprise as well as biotechnology industry clients. She holds a PhD in Public Health Genetics from the University of Washington School of Public Health and was a National Library of Medicine Fellow in Biomedical and Health Informatics in the University of Washington BIME department.
PAPERS & PRESENTATIONS
Zhaoyi Sun, Yujuan Fu, Wen-Wai Yim, Meliha Yetisgen and Fei Xia. An Enhanced Multimodal Multilingual Dataset for Medical Misinformation Detection. Accepted by the 12th IEEE International Conference on Healthcare Informatics (IEEE ICHI 2024).
UPCOMING EXAM
Title: Assessing Disparities Through Missing Race and Ethnicity Data: Results from a Juvenile Arthritis Registry
Student: Katelyn Banschbach
Date/Time: Monday 4/29 at 2:30-3:50
Location: Zoom only – https://washington.zoom.us/my/peter.th
Abstract: Racial and ethnic minorities remain underrepresented in research despite similar willingness to participate. Incomplete race and ethnicity data can lead to exclusion in analysis and those missing this data are more likely to be Black or Hispanic, further worsening disparities. Research and secondary analytics done with incomplete race and ethnicity can unintentionally worsen disparities. Alternatively, missing data may obscure disparities which are already present. Ensuring high quality race and ethnicity data within the EHR and across linked systems, such as patient registries, allows identification of disparities and is necessary to achieve a goal of inclusion of racial and ethnic minorities in scientific research.
Missing race and ethnicity data was assessed and completed within Pediatric Rheumatology Care Outcomes Improvement Network (PR-COIN). The project consists of 4 Aims: (1) Identifying baseline missing race and ethnicity data, (2) Understand current race and ethnicity collection practices and entry into the registry at each center via a REDCap survey, (3) Data completion through three audit and feedback cycles where reports of patients with missing data are sent to each center with request for manual completion via EHR data, (4) Impact assessment on outcome measures via comparison of racial and ethnic differences in risk of certain outcome measures such as elevated clinical juvenile arthritis disease activity score (cJADAS) which are compared pre and post data completion.
The PR-COIN database contains over 5,000 active patients with juvenile idiopathic arthritis spanning 50,000 encounters with plans to add more pediatric rheumatologic diseases over time. Completing missing race and ethnicity data will help avoid unintentionally building inequitable algorithms and system frameworks. We describe the process of identifying and completing missing race and ethnicity data at six centers within the PR-COIN network and highlight the impact of completed data on outcome assessments.
April 1 – April 5, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Aakash Sur, PhD
Thursday, April 11th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: Bioinformatics in Oncology Drug Discovery
Abstract: In the ever-evolving landscape of oncology drug development, biologics have surged in popularity in recent years, and have been adopted as frontline therapy for several indications. Antibody drug conjugates offer a targeted approach to cancer treatment, delivering payloads specifically to cancerous cells and stimulating immune cells instead of relying on the systemic effects of a small molecule. This talk will explore the applications of bioinformatics in early discovery, where potential targets are validated using extensive multi-omic datasets, sequences are confirmed using transcriptome assembly, and antibodies are sequenced with long-read technology. Additionally, as promising payloads and targets are identified, single cell experiments can help elucidate their mechanism of action and understand how the tumor and tumor microenvironment change as a response to therapeutics.
Speaker Bio: Aakash Sur earned a B.S. in Biochemistry and a B.A. in humanities from the University of Texas in 2015 and received a PhD in Biomedical Informatics from the University of Washington in 2022. His thesis work focused on leveraging machine learning to improve the genome assembly of newly sequenced species. Since then, he joined Seagen as a scientist in their research organization with a purview in sequencing related experiments and a focus on single cell technologies. He continues his journey at Pfizer after their acquisition of Seagen in 2023.
PAPERS & PRESENTATIONS
Faisal Yaseen, Rafael Santana-Davila, Clemens Grassberger, Delphine L. Chen, Christina Baik, Keith D. Eaton, Diane Tseng, Smitha Patiyil Menon, Lei Deng, Sylvia Lee, Ariana D Jimenez, Paul D Lampe, A. McGarry Houghton, Paul E Kinahan, Jing Zeng, Stephen R. Bowen. Early FDG PET imaging and circulating T-cell repertoire biomarkers of response to chemotherapy and PD-1 checkpoint inhibitors in patients with stage IV NSCLC. 2024 ASCO Annual Meeting.
ANNOUNCEMENTS
A Population Health Initiative Tier 2 Pilot Grant titled “Culturally adapting and pilot testing chatbot-delivered psychotherapy for Chinese American families caring for older adults with chronic conditions” was funded. The PI is Jingyi Li, and the Co-Investigators are Serena Jinchen Xie, Weichao Yuwen, & Trevor Cohen.
March 25 – March 29, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Laura Wiley, PhD
Thursday, April 4th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: The Importance of Computational Phenotyping in a Learning Healthcare System
Abstract: Every step within the learning healthcare system is dependent upon accurate identification of the patient population of interest. Just as inclusion/exclusion criteria impact the generalizability of randomized controlled trials, the algorithm used for population identification impacts the applicability and generalizability of EHR-based evidence generation. This seminar will discuss my group’s effort to improve the equity and reproducibility of this process to ensure that all patients receive equal benefit from the learning health system.
Speaker Bio:Dr. Laura Wiley is an Associate Professor of Biomedical Informatics at the University of Colorado Anschutz Medical Campus. Her work focuses on computational phenotyping and other informatics methodologies for clinical evidence generation in support of precision medicine. She is an active educator directing the Coursera Clinical Data Science Specialization – a series of 6 MOOCs providing hands on training in clinical research informatics. She is also a national leader in the American Medical Informatics Association having served as Chair or Vice Chair of several AMIA conferences.
PAPERS & PRESENTATIONS
Feng Chen, Liqin Wang, Julie Hong, Jiaqi Jiang, Li Zhou. Unmasking bias in artificial intelligence: a systematic review of bias detection and mitigation strategies in electronic health record-based models. Journal of American Medical Informatics Association. 23 March, 2024. https://doi.org/10.1093/jamia/ocae060.
Siru Liu, Allison B. McCoy, Aileen P. Wright, Scott D. Nelson, Sean S. Huang, Hasan B. Ahmad, Sabrina E. Carro, Jacob Franklin, James Brogan, Adam Wright. Why do users override alerts? Utilizing large language model to summarize comments and optimize clinical decision support. Journal of the American Medical Informatics Association, 2024, 1–9 https://doi.org/10.1093/jamia/ocae041.
F.L. Nkoy, B.L. Stone, Y. Zhang, and G. Luo. A Roadmap for Using Causal Inference and Machine Learning to Personalize Asthma Medication Selection. JMIR Medical Informatics, 2024.
ANNOUNCEMENT
Bhargav Vemuri was selected for the ITHS TL1 Translational Research Training Program’s 2024-2025 cohort for his research proposal on using deep learning to reveal subgroups of developmental trajectories in adolescents. TL1 is a one-year mentored research training program in translational science for predoctoral students.
March 18 – March 22, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Jennifer Hadlock, MD
Thursday, March 28th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: Accelerating research for disease prevention: risk-enriched cohorts for observational studies
Abstract:To accelerate research into disease prevention, it is valuable to conduct prospective studies of the trajectories that begin before disease onset. However, the large cohort size needed can be a deterrent for using advanced methods of observation. This can drive large national studies to use narrow sets of well-established observation modalities. For newer modalities, the cost of large cohorts is a deterrent that drives research toward more reactive, post-diagnosis investigations. Risk-enrichment provides an opportunity to reduce the size of the cohort needed for pre-onset studies of disease trajectories. However, whether using traditional risk scores or new machine learning models, this approach comes with both explicit and potentially hidden trade-offs. Here, we discuss a formal approach for risk-enrichment to reduce cohort size for studies of new-onset diagnosis, and propose approaches for assessing and addressing trade-offs.
Speaker Bio: Dr. Jennifer Hadlock’s research focuses on accelerating translational research into transitions between wellness and disease, by integrating clinical data into systems biology at scale. Her lab develops models from high-fidelity, longitudinal observations of multiomics, phenotype, exposures and patient-reported outcomes. Specific areas of focus are immune-mediated inflammatory disease, maternal-fetal health, and chronic multimorbidity. She is also a PI on the NIH NCATS Biomedical Translator Consortium. She received her MD at the University of Washington School of Medicine, including training in the Rural/Urban Underserved Pathway. Prior to that, she was a Principal Software Engineer in research and development at Microsoft, analyzing and optimizing end-user quality for natural language processing, geographic information systems and real-time digital imaging. She successfully led numerous teams working on technologies used by hundreds of millions of people worldwide.
PAPERS & PRESENTATIONS
Yue Guo, Joseph Chee Chang, Maria Antoniak, Erin Bransom, Trevor Cohen, Lucy Lu Wang, Tal August. Personalized Jargon Identification for Enhanced Interdisciplinary Communication. Accepted by NAACL 2024.
ANNOUNCEMENTS
Ehsan Alipour was selected as a recipient of this year’s Reviewer Award for the AMIA Informatics Summit. Nice work!
Anne Turner will be presenting a webinar entitled “Decision-Making in Dementia Care: Preferences of People with Memory Loss” at the PennAITech Collaboratory for Healthy Aging Webinar Series on April 4th. To register, please click here.
March 11 – March 15, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590 – On break until March 28th!
PAPERS & PRESENTATIONS
Yujuan Fu*, Giridhar Kaushik Ramachandran*, Nicholas J Dobbins, Namu Park, Michael Leu, Abby R. Rosenberg, Kevin Lybarger, Fei Xia, ̈Özlem Uzuner, and Meliha Yetisgen. Extracting social determinants of health from pediatric patient notes using large language models: Novel corpus and methods. Accepted by LREC-COLING, 2024.
Wen-wai Yim, Yujuan Fu, Asma Ben Abacha, and Meliha Yetisgen. To err is human, how about medical large language models? comparing pre-trained language models for medical assessment errors and reliability. Accepted by LREC-COLING, 2024
Namu Park, Kevin Lybarger, Giridhar Kaushik Ramachandran, Spencer Lewis, Aashka Damani, Özlem Uzuner, Martin Gunn and Meliha Yetisgen. A Novel Corpus of Annotated Medical Imaging Reports and Information Extraction Results Using BERT-based Language Models. Accepted for 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024).
March 4 – March 8, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590 – On break until March 28th!
PAPERS & PRESENTATIONS
Zeng, N.C. Jani, A.M. Sotolongo, G. Luo, M. Arjomandi, M.J. Falvo. Pulmonary function, chronic respiratory symptoms, and functional limitation among Veterans in the Airborne Hazards and Open Burn Pit Registry. American Thoracic Society International Conference, San Diego, CA, May 2024.
GENERAL EXAM
Title: Use of the Electronic Health Records to facilitate phenotyping, comorbidity analysis, and genomics
Student: Su Xian
Date/Time: 3:30 pm, March 7th, 2024 – TODAY!
Location – Zoom only: http://bit.ly/seanmooneyzoom
Abstract: One of the contemporary medical research challenges is EHR-based digital phenotyping. The electronic MEdical Records and Genomics (eMERGE) consortium has launched Phenotype KnowledgeBase (PheKB), an EHR-based Phenotyping knowledge base engaging multiple sites of large hospitals and universities to share their phenotyping algorithms developed using the EHR data. Though most of the algorithms stored in PheKB are rule-based and validated by domain expertise, there are efforts to develop machine learning algorithms for EHR-based phenotyping. Various types of machine learning and deep learning methods have been tested to derive EHR-based phenotyping algorithms, including Support Vector Machines (SVM), random forest, logistic regressions, and neural network architectures. In this work, we will: 1) Explore the potential of using EHR data to phenotype patients. We will develop an unsupervised machine learning algorithm, starting from learning the diagnosis codes and procedure codes, to finally learning numerical embeddings for each patient, using the data from the eMERGE consortium. 2) Evaluating the performance of the algorithm on disease prediction and bulk phenotyping. Applying the designed algorithm to study the comorbidities of phenotypes, revealing heterogeneity using a few selected phenotypes as cases, including colorectal cancer, and systemic lupus erythematosus, to facilitate personalized treatment and gain an understanding of the complicated disease. 3) Using a rule-based phenotyping algorithm derived from the eMERGE consortium, we focus on a complicated psychological disease – depression – using EHR-based phenotyping algorithms to study the genetic risk factors and uncover the molecular mechanisms of depression.
ANNOUNCEMENTS
Nic Dobbins, PhD has accepted a position as an Assistant Professor in the Johns Hopkins Dept. of Medicine, in the Biomedical Informatics & Data Science section, starting 4/1. Congrats Nic!
From Savitha Sangameswaran: Invitation to Participate in Research Study of Information in Translational Research Teams
Are you a clinical and translational researcher? You are invited to participate in a study on how funded translational research teams manage information. This study Information Management Prototype for Clinical and Translational Research (IMPACT-CTR), is funded by an RO1 from National Library of Medicine and aims to understand the tools and strategies teams use in seeking, using, creating, sharing, storing, and retrieving information while conducting collaborative clinical and translational research. We will use what we learn from the study to create training materials to help teams develop evidence-based information strategies that can make CTR more efficient and effective.
Please visit https://impactctr.sagebionetworks.org/ for more information about this study and how you can participate!
To participate in this study, please complete this interest form and a member of the research team will reach out to schedule a brief informational call!
February 26 – March 1, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Diane M. Korngiebel, DPhil, MA
Thursday, March 7th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: Anthropomorphism: What is it and why does it matter for Artificial Intelligence applications?
Abstract: Dr. Korngiebel will present on the importance of considering anthropomorphism in the context of generative AI applications that leverage Large Language Models (LLMs) and Large Multi-model Models (LMMs). The presentation will start by defining key terms, discuss anthropomorphism and how it relates to generative AI, and suggest areas for further research into both design and deployment that take into account the challenges and opportunities that accompany anthropomorphism.
Speaker Bio: Diane M. Korngiebel is currently a Senior AI Ethicist in Trust & Safety at Google and has been an ELSI (ethical, legal, and social implications) Scholar at Google since Oct. 2021. Before joining Google, Dr. Korngiebel spent a year at The Hastings Center as a Research Scholar, and until 2020, she was an Associate Professor in the Department of Biomedical Informatics and Medical Education and an adjunct Associate Professor in the Department of Bioethics and Humanities at the University of Washington School of Medicine in Seattle; she maintains affiliate faculty status in both UW departments. Her current interests focus on the ethics of using generative AI in healthcare and wellness support.
Dr. Korngiebel’s work has appeared in the American Journal of Public Health, Nature: Genetics in Medicine, NPJ Digital Medicine, and PLoS Genetics. Before joining Google, Dr. Korngiebel was the Principal Investigator on three NIH ELSI program grants. She is the chair of the AMIA ELSI Working Group, serves as an IEEE working group member developing recommendations for organizations regarding AI governance, and has helped inform ISO standards for health-related smartphone apps.
Medical Data Science Seminar – NEXT Tuesday, March 5th at 1pm PT – Zoom link
Creating a unique Acute Myeloid Leukemia (AML) Digital Twin for cancer patients contributes to innovative treatments
Ilya Shmulevich, PhD
Professor, Institute for Systems Biology & Affiliate Professor
Departments of Bioengineering and Electrical Engineering, UW
PAPERS & PRESENTATIONS
Nikita Pozdeyev, Manjiri Dighe, Martin Barrio, Christopher Raeburn, Harry Smith, Matthew Fisher, Sameer Chavan, Nicholas Rafaels, Jonathan A Shortt, Meng Lin, Michael G Leu, Toshimasa Clark, Carrie Marshall, Bryan R Haugen, Devika Subramanian, Kristy Crooks, Christopher Gignoux, Trevor Cohen. Thyroid Cancer Polygenic Risk Score Improves Classification of Thyroid Nodules as Benign or Malignant. The Journal of Clinical Endocrinology and Metabolism. 2024 Jan;109(2):402-412. DOI: 10.1210/clinem/dgad530. PMID: 37683082.
Link to featured article: https://academic.oup.com/jcem
GENERAL EXAM
Title: Operationalizing Predictive Decision Support and Disease Phenotyping to Improve Healthcare Outcomes in Chronic Obstructive Pulmonary Disease
Student: Siyang (Sunny) Zeng
Time: 7:00 am, March 1st, 2024
Location: Zoom only – https://washington.zoom.us/j/4666998448?pwd=dXo1NjFCQkNJclFYc2Y0SHN3c0JPZz09
Meeting ID: 466 699 8448, Password: 524369)
Abstract:
Chronic obstructive pulmonary disease (COPD) is a major cause of death and places a heavy burden on healthcare. Severe COPD exacerbations requiring emergency department visits or inpatient stays often cause irreversible decline in lung function and health status and account for majority of the total medical cost related to COPD. Successful preventive strategies such as care management can reduce severe exacerbations in patients with COPD, however, due to limited resources and service capacity, only a small portion of patients could enter a care management program. Thus, its effectiveness is upper bounded by how accurately it enrolls patients who are at risk for severe exacerbations. There is a need to facilitate accurate pinpointing of high-risk patients in order to optimize resource allocation and patient outcomes.
Although predictive models can identify high-risk patients for preventive care, there are several obstacles to operationalize the predictive decision support and improve outcomes in practice: (1) existing predictive models for severe COPD exacerbations are inaccurate and suboptimal for clinical use; (2) predictive models lack explanability and interpretability, forming barriers for making sense of the prediction or identifying the most effective preventive strategies; (3) despite being built with the purpose of forewarning to prevent future exacerbations, existing predictive models were seldom evaluated for their impact on actual clinical actions; and (4) incomplete linkages between disease mechanisms and clinical outcomes hinders the effect of therapies or management strategies.
In this research, we aim to tackle these obstacles by (1) developing the most accurate model to predict patients at high-risk for future severe COPD exacerbations, (2) providing explanations with clinical recommendations for the high-risk prediction, (3) piloting a user study to evaluate the impact of providing explanations with clinical recommendations for a high-risk prediction on care management enrollment decisions, and (4) determining lung function phenotypes associated with COPD mechanisms and clinical outcomes.
February 19 – February 23, 2024
UPCOMING LECTURES AND SEMINARS
BIME 590
Sarah Biber, PhD
Thursday, February 29th – 11-11:50 am
850 Republican Street, Building C, Room 123 A/B
Zoom Information: https://washington.zoom.us/my/bime590
Title: A Next Generation Multimodal Data Platform for Discovery and Translation in Alzheimer’s
Abstract: Alzheimer’s is a complex and devastating disease. The advent of breakthrough disease modifying therapeutics like Leqembi have created new hope for a future where Alzheimer’s disease can be detected and treated before cognitive symptoms occur. However, developing effective early detection tools, gaining a comprehensive understanding of disease drivers and mitigators, and developing precision therapeutics requires a holistic understanding of the disease. This ambitious task demands data driven approaches and leveraging disparate modalities of research and clinical data that are harmonized and shareable. The National Alzheimer’s Coordinating Center has built a next generation multimodal data integration and harmonization platform that will enable researchers to ask and answer the most pressing questions in the field and gain critical new insights into Alzheimer’s.
Presenter Bio: Sarah Biber, PhD, is the Executive Director of the National Alzheimer’s Coordinating Center (NACC), based at the University of Washington. She co-leads NACC’s scientific and strategic direction and $58M project portfolio, and functions as a co-PI with NACC’s PI and Director, Dr. Walter Kukull. Within this role, Dr. Biber represents NACC with national and international partners, spearheads national initiatives, leads development of major grant applications, leads academic and industry partnerships, and oversees NACC’s tech, operations, research, communications, and grants and finance teams. Under her leadership, NACC has undergone a massive informatics transformation to establish a next generation multimodal data platform for discovery and translation in Alzheimer’s Disease and Related Dementias (ADRD).
PAPERS & PRESENTATIONS
Payne, Thomas. I am not burned out. This is how I write notes. Accepted abstract, AMIA Clinical Informatics Conference, May 2024.