Khaleghzadegan S, Rosen M, Links A, et al. Validating computer-generated measures of linguistic style matching and accommodation in patient-clinician communication. Patient Educ Couns. 2024;119:108074. doi:10.1016/j.pec.2023.108074
This analysis of previously collected data from the cross-sectional Enhancing Communication and HIV Outcomes (ECHO) study examined aspects of linguistic style matching (LSM) between patients and clinicians. The ECHO study was designed to explore racial and ethnic disparities in patient-clinician communication in HIV care. Each patient had one audio-recorded session with their primary HIV clinician that was analyzed with their consent. The analysis outlined in this paper examined the transcripts from these encounters. A total of 330 transcripts were assessed using four computer-generated measures of LSM and linguistic accommodation (LA) to explore the degree to which clinicians adapt to their patients’ linguistic style throughout a visit. Patients were predominantly Black and male, while clinicians were mostly White and female. Generally, the developed LSM metrics and patient-reported clinician speech similarity were not significant. Future research may address these differences in perception and whether LSM or patient-perceived similarity enhances patient outcomes. In dynamics where LSM was low at the start of the session, LA was higher, and indicated, given the trajectory of LA and patient-rated speech similarity, a successful accommodation. Conversation-level LSM was associated with less clinician verbal dominance, indicating higher patient-centered communication. Two measures (rolling-window reciprocal LSM and reciprocal LSM) showed negative associations with speech similarity and positive associations with clinician dominance. While this may be due to nuances in clinician-dominated encounters, further research is needed to explore these results. Overall, these data support the need for additional understanding of LSM (and its parts) as well as LA with diverse patients and clinicians.