(HealthDay News) — Writing style can explain about half of observed disparities in care team response to messages sent by patients, according to a study published online Aug. 24 in JAMA Network Open.
Mitchell Tang, Ph.D., from Harvard University in Cambridge, Massachusetts, and colleagues quantified factors that could potentially explain lower rates of care team response to messages sent by patients from historically marginalized groups in a cross-sectional study using natural language processing. Message content and writing style features were extracted from portal message threads.
The sample included 3,619,390 patient message threads sent by 511,020 patients. Of the message threads, 4.0, 3.7, 5.9, 83.0, 1.4, and 2.1 percent were sent by patients who were Asian, Black, Hispanic, White, other race and ethnicity, and unknown race and ethnicity, respectively. The researchers found that the levels of care team response were lower for messages sent by patients from historically marginalized groups. Black patients had a 3.7 percentage point lower response rate from the target clinician compared with White patients, representing an 11.6 percent reduction. There was an association for message content with target clinician response, but it could not explain these differences. Writing style accounted for 48.0, 34.9, 60.5, and 42.8 percent of the difference for Black patients, Hispanic patients, patients with only a high school education (versus a college degree), and Medicaid patients (versus commercially insured patients).
“The findings could help health systems identify potential sources of bias in portal-message triage and develop more equitable approaches to digital communication,” Tang said in a statement.
