Abstract
National COVID Cohort Collaborative (N3C) enclave provides health researchers with a rich dataset from 76 contributing clinical sites. However, the harmonized data lacks certain details available in sites' local electronic health records (EHRs), such as the principal diagnosis code for reported emergency department (ED) and inpatient (IP) visits. This means a principal diagnosis of COVID-19 can only be inferred by applying a time relationship between the visit dates and the record of infection and diagnosis. The purpose of this study is to perform a single-site sensitivity analysis modeled after an N3C study examining potential race-ethnicity based bias in hospitalization decisions during COVID-19 related ED visits. The analytic pipeline was first run in N3C, then reproduced locally with N3C data fields from a single-site, and finally run a third time using the additional principal diagnosis data. We find the effects of patient comorbidities and race-ethnicity groups on direct IP admittance to be consistent among the three cohorts with varying levels of statistical significance due to different sample sizes.
Original language | English (US) |
---|---|
Title of host publication | 2023 Systems and Information Engineering Design Symposium, SIEDS 2023 |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 297-302 |
Number of pages | 6 |
ISBN (Electronic) | 9798350300642 |
DOIs | |
State | Published - 2023 |
Event | 2023 Systems and Information Engineering Design Symposium, SIEDS 2023 - Charlottesville, United States Duration: Apr 27 2023 → Apr 28 2023 |
Publication series
Name | 2023 Systems and Information Engineering Design Symposium, SIEDS 2023 |
---|
Conference
Conference | 2023 Systems and Information Engineering Design Symposium, SIEDS 2023 |
---|---|
Country/Territory | United States |
City | Charlottesville |
Period | 4/27/23 → 4/28/23 |
Bibliographical note
Publisher Copyright:© 2023 IEEE.