141 Frequency and Effect of Interruptions on Resident Workload in the Emergency Department

Abstract

Quantify the frequency of interruptions and effect on resident cognitive workload in an academic emergency department (ED). Emergency medicine (EM) residents of varying post-graduate year (PGY) were directly observed during 8-hour clinical ED shifts by experienced health systems researchers. Interruptions were classified by type (face-to-face, phone call, page, environmental, patient care demand, technology), priority (high, medium, low), and location (eg, staff work station, patient room) using a validated tablet PC-based tool. Residents completed subjective assessments of mental demand task complexity, temporal demand, distractions, situational stress, and physical demand using the NASA-Task Load Index (TLX) and reaction time tests at the mid- and end of shifts. Twenty-three resident shifts (PGY1=8, PGY2=8, PGY3=7) were observed in the ED. There were 2132 interruptions (M=11.6 interruptions/hour, SD=3.29). The average duration of an interruption was 37.5 seconds. More than 12% of clinical time per shift was spent on interruptions (3.5% high priority, 7.8% medium priority, 1.3% low priority). Face-to-face nurse communication (24.7%), face-to-face physician communication (21.4%), and face-to-face other communication (12.9%) interruptions were the most prevalent interruption types experienced by residents. Approximately three-quarters (77.9%) of interruptions were of medium priority, followed by low priority (13.6%) and high priority (8.5%). The majority (83.7%) of interruptions occurred in the staff work station and 10.7% in the patient room. There was a significant difference in the number of interruptions experienced by PGY level (F(2,20) =5.4, p=0.01). EM PGY3s faced more interruptions (M=14.4 interruptions per hour, SD=2.9) than PGY2s (M=11.0, SD=2.59) or PGY1s (M=9.8, SD=2.83). NASA-TLX scores were universally higher at the end of the shift across all 6 dimensions (Figure 1); however, significant increases were only observed for mental demand (p=0.02) and physical demand (p=0.01). Reaction times were not significantly higher at the end of the shift compared to the beginning of the shift (p = 0.34). No significant differences were observed in performance scores or reaction times with different interruption loads per shift. Residents experienced more interruptions than reported in the literature. Nearly one hour of residents’ clinical time was spent dealing with interruptions. The majority of interruptions were face-to-face communication with other health care team members, of medium priority, and occurred at the staff work station. PGY3 EM residents faced significantly more interruptions than their counterparts, which could be reflective of increased responsibility as residents gain seniority. Overall, residents revealed increasing cognitive demand and slower reaction times over the course of the shift. Higher powered studies may be needed to detect differences in performance scores and reaction times with different interruption loads. Further work is needed to assess the positive and negative implications of interruptions on resident workload and what interventions will be most effective to minimize interruptions that occur in the ED workspace.