# Examining the relationships among physician implicit bias, language, and Hispanic patient satisfaction

## TL;DR

This study found that higher levels of a doctor's hidden anti-Hispanic bias are linked to lower patient satisfaction, especially when bias is high, and language used may also play a role.

## Contribution

The study reveals a quadratic relationship between physician implicit anti-Hispanic bias and patient satisfaction, and suggests language variables may mediate this relationship.

## Key findings

- Higher implicit anti-Hispanic bias in physicians is associated with lower patient satisfaction at high bias levels.
- Language variables interact with physician bias to predict patient satisfaction, indicating context matters in bias communication.

## Abstract

The present study examined links among physician implicit bias and physician word use and patient satisfaction during outpatient medical visits.

To test these relationships, we measured implicit anti-Hispanic bias of 53 internal and family medicine residents and audio recorded outpatient visits between these residents and 291 of their Hispanic-identifying patients between 2015 and 2017. After each visit, both patients and resident physicians completed surveys measuring their perceptions of the interactions (residents) or their satisfaction with their care (patients). Linguistic Inquiry and Word Count was used to quantify resident word use from each audio recording.

A quadratic relationship between physician implicit bias and patient satisfaction suggested that increased levels of implicit anti-Hispanic bias were related to lower levels of patient satisfaction, but only at high levels of physician bias. Exploratory analyses revealed that several language variables interacted with physician implicit bias to predict patient satisfaction suggesting that implicit bias may also be communicated by more moderately biased physicians in some contexts.

These results add to a growing literature establishing links between physician implicit bias and patient care and suggest that future work should explore how context impacts the communication of physician bias to patients.