Banner

News

Article

Does racial or ethnic bias guide responses to EHR portal messages?

Researchers examine responses from physicians and nurses to patients in minoritized groups.

medical technology ehr: © tippapatt - stock.adobe.com

© tippapatt - stock.adobe.com

Attending physicians were more likely to respond to electronic health record (EHR) portal messages from White patients than from minoritized patients, according to a new study.

However, care team response rates were relatively even when comparing White, Black, Hispanic or Asian patients, according to researchers.

By now it’s no secret that as the COVID-19 pandemic grew, so did use of EHR portals as patients sent secure messages to their physicians, and received responses the same way, according to the original investigation, “Differences in Care Team Response to Patient Portal Messages by Patient Race and Ethnicity,” published in JAMA Network Open.

“Higher message volumes have largely persisted, reflecting a new normal, with portal messaging as a central channel for patient–health care professional interaction,” the study said. With physicians at times overwhelmed with patient queries, triaging nurses may respond themselves or forward them to other health care professionals, including doctors.

Numerous messages

Researchers examined who responded to 39,043 primary care patients at Boston Medical Center in 2021. They found “relatively small differences in overall care team response rates” across the four racial and ethnic groups in the sample.

For example, the overall response rate was 65.7%. By racial and ethnic group, response rate was 70.9% to Asian patients, 66.6% to Hispanic patients, 66.3% to Black patients, and 65.7% to White patients.

However, White patients were more likely to get a reply from the attending physician – 22.7%, compared with 22.1% to Asian patients, 18.1% to Hispanic patients, and 16.7% to Black patients.

The lower response rates could be due to:

  • Differences in the underlying message request, or the issue and what the patient was asking about.
  • Differences in communication, accounting for specificity, tone, grammar and spelling.
  • Implicit bias, conditioned on the first two characteristics.

The researchers noted the differences persisted even when accounting for the first two factors.

As for bias, “within the context of portal messaging, racial bias may manifest in lower likelihood of care team response for messages from patients who belong to minoritized racial and ethnic groups,” the study said.

“With nurse triaging, there is the specific concern that messages from patients who belong to minoritized racial and ethnic groups are less likely to be prioritized for physician response,” the study said.

Black patients had the highest response rate by registered nurses, who answered back in 34.1% of messages, compared with 32.1% for Asian patients, 32.2% for Hispanic patients, and 22.7% of White patients, according to the study. Response percentages were lower for advanced practice professionals (average 7.1%), any resident (average 2.6%), and any other response (12%).

Appropriate responses?

The researchers said they were unaware of any prior work that has examined benefits of having an attending physician or a registered nurse respond. “Indeed, for many types of message requests, direct RN response may be most appropriate,” the study said.

Health literacy also may play a role. The researchers noted earlier studies found patients in minoritized racial and ethnic groups have lower health literacy, and that could influence the types of requests and how those are communicated in the EHR portals.

If racially and ethnically minoritized patients are less aware of the types of messages that can be sent through patient portals, there may be solutions to help patients close the response gap. For example, educational videos on portal use could be shown in waiting rooms or there could be primers and tutorials in the portal itself. Portal modifications using artificial intelligence also could help patients better categorize and structure messages, the study said.

Related Videos
© drrobertkushner.com
© drrobertkushner.com
© drrobertkushner.com
© drrobertkushner.com