Collection and Validation of Patient Self-Reported Race, Ethnicity, and Language (REL) Information In a Postpartum Setting
Date
2023-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
As healthcare organizations move toward accountable care agreements, and move
away from fee-for-service, there is a greater need for healthcare organizations to have
stronger data to support population-based interventions. Moreover, literature has
highlighted the discrepancies in collecting identity-based information from patients,
including information regarding patient race, ethnicity, and language (REL). REL data
has multiple policy and clinical implications as it is utilized to not only determine the
allocation of funds for programs but also is used to create evidence-based
interventions to decrease health disparities. If the core of this data is incorrect, then
resource allocation is futile. More importantly, there is a potential that the resources
and interventions that are being created using this data are now not effectively
reaching and impacting these communities. Prior research demonstrates that there are
large disparities in women’s health, especially by race.1 Given what we know about
the nature of flawed data, these disparities are potentially increased or inadequately
captured by current interventions. In an effort to assess organizational capacity to
collect REL data and identify where discrepancies in the documentation of REL data
may occur, this quality and safety improvement project assesses the practices of
collecting REL data from patients, as well as concurrence or discrepancy in how REL
data is documented within the patient’s chart and how they choose to self-identify.
These concurrences and discrepancies were measured with a two-pronged approach
where one prong involved patient survey of self-identified REL information and the
second involved collection and validation from the Electronic Health Record (EHR).
Study results demonstrated an overall concordance between the two data corpuses;
however, the discrepancies and variation in certain minority groups were noteworthy.
Given the results, the main finding is the need for an EHR with broader fields and/or
allowing patients to self-identify their demographic data to allow for the validation of
patient identities, create accurate data corpuses, and improve patient health outcomes.
Once modified, we expect researchers to have more accurate and credible data to
identify health disparities from, driving the eventual closure of the inequities seen
within multiple minority populations.