Abstract Number:
1281
Submission Type:
Invited Paper Session
Participants:
Zheyu Wang (1), Zheyu Wang (1), Xu Shi (2), Yuxin Zhu (1), Aaron Miller (3), Rebecca Hubbard (4)
Institutions:
(1) Johns Hopkins University, N/A, (2) University of Michigan, N/A, (3) University of Iowa, N/A, (4) University of Pennsylvania, N/A
Chair:
Session Organizer:
Speaker(s):
Session Description:
This session is organized to echo the overarching theme of JSM 2024: "Statistics and Data Science: Informing Policy and Countering Misinformation". In the current data-rich landscape, Electronic Health Records (EHRs) play a pivotal role in advancing healthcare practice and policy. However, their effective utilization demands careful consideration and expertise.
On the front end, with a multitude of data partners and medical coding systems, the variability in coding clinical concepts is a prevalent concern. Even with the implementation of Common Data Models designed to standardize data elements, differences in coding "dialects" remain due to care practices, financial drivers, and data-sharing constraints. To this end, Dr. Xu Shi (U Michigan) will present data-driven and privacy-preserving statistical methods for detecting and reducing coding differences between healthcare systems (Title: Harmonizing Electronic Health Record and Claims Data Across FDA Sentinel Initiative Data Partners: Case Study and Lessons Learned).
On the back end, properly harmonized EHR data offer great opportunities to inform policy and counter misinformation. One significant area of exploration is to study the characteristics of current healthcare practices and identify areas for improvement. For example, how to quantify and monitor misdiagnosis-related harm in existing practice to offer insights into potentially preventable harm? This session invites speakers to present from two different angles. The first adopts a population-level approach with minimal assumption, intended to inform policy and healthcare practices from a broader perspective (Dr. Yuxin Zhu, Johns Hopkins) (Title: Quantifying Misdiagnosis-related harm leveraging health record data and through mixture-model-based novel measures). Alternatively, the second speaker on this topic adopts a more individualized perspective by employing different modeling structures to investigate cases at a granular level. And therefore enables an in-depth exploration of factors contributing to diagnostic delays, paving the way for potentially personalized healthcare solutions (Dr. Aaron Miller, U Iowa) (Title: Bootstrapping-based approaches to estimate the frequency, duration and risk factors for diagnostic delays). Another example is utilizing EHR to improve personalized medicine by targeting interventions to those individuals most likely to experience benefits. However, such effort may involve issues associated with risk model-guided medical practice, particularly concerning underrepresented race and ethnicity groups. Dr. Rebecca Hubbard (U Penn) will conclude the session with such a discussion. She will illustrate this point with an example of risk-targeted breast cancer screening and propose alternative approaches with EHR data to minimize these biases (Title: Statistical challenges in development of equitable approaches to risk-guided cancer screening using EHR data).
All speakers have agreed to attend JSM if session is selected, and none of them has indicated conflicts with specific meeting dates.
Sponsors:
Biometrics Section 1
Health Policy Statistics Section 2
Society for Medical Decision Making 3
Theme:
Statistics and Data Science: Informing Policy and Countering Misinformation
Yes
Applied
Yes
Estimated Audience Size
Small (<80)
I have read and understand that JSM participants must abide by the Participant Guidelines.
Yes
I understand and have communicated to my proposed speakers that JSM participants must register and pay the appropriate registration fee by June 1, 2024. The registration fee is nonrefundable.
I understand