Characterizing Multitasking and Workflow Fragmentation in Electronic Health Records among Emergency Department Clinicians: Using Time-Motion Data to Understand Documentation Burden

Abstract

Background: The impact of electronic health records (EHRs) in the emergency department (ED) remains mixed. Dynamic and unpredictable, the ED is highly vulnerable to workflow interruptions.

Objectives: The aim of the study is to understand multitasking and task fragmentation in the clinical workflow among ED clinicians using clinical information systems (CIS) through time-motion study (TMS) data, and inform their applications to more robust and generalizable measures of CIS-related documentation burden.

Methods: Using TMS data collected among 15 clinicians in the ED, we investigated the role of documentation burden, multitasking (i.e., performing physical and communication tasks concurrently), and workflow fragmentation in the ED. We focused on CIS-related tasks, including EHRs.

Results: We captured 5,061 tasks and 877 communications in 741 locations within the ED. Of the 58.7 total hours observed, 44.7% were spent on CIS-related tasks; nearly all CIS-related tasks focused on data-viewing and data-entering. Over one-fifth of CIS-related task time was spent on multitasking. The mean average duration among multitasked CIS-related tasks was shorter than non-multitasked CIS-related tasks (20.7 s vs. 30.1 s). Clinicians experienced 1.4 ± 0.9 task switches/min, which increased by one-third when multitasking. Although multitasking was associated with a significant increase in the average duration among data-entering tasks, there was no significant effect on data-viewing tasks. When engaged in CIS-related task switches, clinicians were more likely to return to the same CIS-related task at higher proportions while multitasking versus not multitasking.

Conclusion: Multitasking and workflow fragmentation may play a significant role in EHR documentation among ED clinicians, particularly among data-entering tasks. Understanding where and when multitasking and workflow fragmentation occurs is a crucial step to assessing potentially burdensome clinician tasks and mitigating risks to patient safety. These findings may guide future research on developing more scalable and generalizable measures of CIS-related documentation burden that do not necessitate direct observation techniques (e.g., EHR log files).

Fig. 1

Distribution of observation days and start times in the emergency department.

Fig. 2

Fictional workflow demonstrating the alignment and calculation of the duration of multitasking between physical and communication task sequences based on clock time. The figure above, depicts three physical and three communication tasks at baseline which, combined, generate three non-multitasked and four multitasked tasks. M, multitasked; NM, not multitasked.

Fig. 3

Task type stratified by physical location of clinician when task was performed.

Fig. 4

Top 80% multitasked tasks stratified by non-CIS and CIS-related tasks. CIS, clinical information system.

Fig. 5

Heatmaps of CIS-related tasks and its preceding or following task stratified by the presence of multitasking. CIS, clinical information system.