Evaluation of Attention Switching and Duration of Electronic Inbox Work Among Primary Care Physicians.
Level 4 - case-series / case-control
Cross-sectional observational study of electronic health record access logs.
PubMed 33475754 · doi:10.1001/jamanetworkopen.2020.31856
What was done
This cross-sectional study analyzed electronic health record (EHR) access logs from March 1 to March 31, 2018, across 1,275 primary care physicians (PCPs) in an integrated health system serving 4.5 million patients. The authors quantified attention switching (defined as switching between electronic inbox tasks, other EHR activities, and non-EHR periods) and daily inbox work duration (minutes spent on message views and related tasks over 24 hours on workdays). Multivariable regression models were used to evaluate factors associated with switching frequency and work duration.
What was found
PCPs received a mean (SD) of 332.6 (148.3) new inbox messages weekly. On workdays, PCPs averaged 79.4 (SD 21.8) attention switches related to inbox work and spent a mean of 64.2 (SD 18.7) minutes on inbox tasks. In multivariable models, each additional message beyond reference levels was associated with increased attention switches: patient secure messages (+0.289 switches; 95% CI, 0.217–0.362), administrative messages (+0.262; 95% CI, 0.166–0.358), results messages (+0.203; 95% CI, 0.127–0.278), and request messages (+0.190; 95% CI, 0.124–0.257). Each additional minute of inbox work was associated with 0.468 switches (95% CI, 0.411–0.524). For inbox duration, each additional results message added 0.338 minutes (95% CI, 0.272–0.404), administrative message added 0.179 minutes (95% CI, 0.093–0.265), patient message added 0.151 minutes (95% CI, 0.085–0.217), and request message added 0.101 minutes (95% CI, 0.041–0.161). Each attention switch was independently associated with an additional 0.373 minutes of inbox work duration (95% CI, 0.328–0.419).
Why it matters
This study objectively quantifies EHR-induced cognitive fragmentation in primary care, demonstrating that incoming message volume is the primary driver of both workflow interruptions and prolonged inbox time, whereas patient panel and physician characteristics play smaller roles.
Limits
The cross-sectional design cannot establish causality. Data were gathered from a single integrated health system during one month in 2018, which may limit generalizability to independent practices or alternative EHR systems. Audit logs infer attention switching indirectly and cannot capture in-person clinical interruptions, cognitive fatigue, or clinical error rates.
Cited by
- context Research by Gloria Mark demonstrates that frequent external distractions habituate the brain to interruptions, causing individuals to self-distract even in the absence of external interruptions.