Ishani Ganguli is a primary care physician, health services researcher and assistant professor of medicine at Harvard Medical School and Brigham and Women’s Hospital and a member of the 2021 Tradeoffs Research Council. She studies medical decision-making and delivery and payment innovation in ambulatory care.

In between patient visits and meetings, primary care clinicians (PCPs) are usually found working through their electronic health record (EHR) inboxes. There’s a message from Phillip who has a new, persistent headache. A pharmacy request from Tina. A CT scan result for Rodrigo. A note from Tabitha’s emergency department visit. Each message requires attention and often some kind of follow-up.

Studies in medicine and other fields show that frequently switching between tasks can be mentally taxing to the point of causing errors, but no one has looked at how much EHRs contribute to this problematic multitasking. So a group of researchers from Kaiser Permanente tracked how 1,275 full-time or near-full-time PCPs used their EHR inboxes over a typical month in 2018 and published their results in JAMA Network Open.*

They found that each doctor received, on average, 333 new messages in their inbox per week. Thirty percent of these were test results; 28% were requests from call centers, pharmacies, or other clinicians; 23% were messages initiated by patients; and 20% were administrative. PCPs switched their attention an average of 79 times between the inbox itself, other parts of the EHR and work outside the EHR.

The researchers found some intriguing associations with these metrics. For instance, male doctors spent two fewer minutes on their inbox yet switched attention three more times per day compared to female doctors. Doctors with a larger share of elderly patients switched attention more often. Yet the most consistent predictor was simply the volume of messages a PCP received — the more messages they got, the longer they spent in their inbox and the more times they switched in and out of it.

This study made clever use of EHR audit data which vendors are required to collect for security purposes and are increasingly repurposed to examine clinicians’ work patterns. In this case, the researchers could track activity in the inbox and other parts of the EHR, and they assumed that if more than two minutes went by without any action, then the PCP was doing work outside the EHR.

The researchers couldn’t tell if the attention switches were from external interruptions or deliberate choices by the physicians, nor whether these switches were linked to worse patient outcomes. But quantifying them is an important step — we’re not going back to paper charts any time soon, and there’s evidence that patients benefit from convenient, portal-based access to their care teams. So how do we ensure that frequent attention switching by PCPs doesn’t contribute to clinician burnout and compromised care?

Health systems and practices can enable PCPs to share the load —whether by divvying up inbox work between clinical team members to the top of each team member’s license or by using artificial intelligence to triage incoming messages.  And we should make sure that physician compensation — still largely linked to office visits and procedures — makes space for this sort of work.

*Ishani Ganguli became an Associate Editor at JAMA Network Open in January 2021. She had no involvement in the editing or publication of this manuscript.