To examine differences in the use of negative descriptors about patients with sickle cell disease (SCD) in electronic health records compared to patients with intersecting stigmatizing identities, such as race, chronic pain, and opioid use disorder.
Key Findings:
Patients with SCD were characterized with negative descriptors more frequently than non-Black patients without chronic pain or OUD, indicating a potential bias in clinical assessments.
Negative descriptors included terms like aggressive, agitated, and angry, which may influence treatment decisions.
Clinician biases were evident in the language used to describe patients with SCD, potentially impacting their care and leading to inequitable treatment.
Interpretation:
The study highlights the presence of clinician biases in electronic health records, particularly against patients with SCD, which may contribute to inequitable treatment and worse health outcomes, necessitating further investigation.
Limitations:
Retrospective nature of the study may introduce biases.
Sample limited to a single institution, affecting generalizability.
Exclusion of notes from other health professionals may limit the comprehensiveness of data.
Potential biases in the natural language processing methodology could affect descriptor identification.
Conclusion:
The findings underscore the need for awareness and training to mitigate biases in clinical language and improve care for patients with SCD.
