Implementing System Improvement Actions After Peer Review of Emergency Department Incident Reports: A...

Emergency departments are high-risk, high-workload settings where safety incidents and near misses may occur. However, incident reporting systems do not consistently translate reported events into timely learning and actionable system change. We evaluated a quality improvement initiative that integrated frontline peer review and recurring joint incident meetings into a centrally managed hospital incident reporting system.

We conducted a single-center retrospective before-and-after evaluation in an urban tertiary-care emergency department. The intervention comprised an introductory session emphasizing a learning-oriented, non-punitive approach to incident reporting; end-of-shift mini-discussions; monthly peer review of incident reports by an emergency department peer-review team; and monthly joint incident meetings with the hospital patient safety management department. We compared 12-month pre-intervention and post-intervention periods separated by a one-month wash-in period. The primary outcome was the number of system improvement actions reaching each of three prespecified cumulative milestones: proposal, approval for implementation, and initiation of implementation as documented in meeting minutes. Actions were also classified as emergency department-only or hospital-wide in scope. Secondary outcomes were incident-reporting volume and degree of harm. Event type categories were described as report characteristics. Monthly reporting rates were analyzed using interrupted time-series regression.

Incident analysis generated 24 proposed system improvement actions; 20 were approved for implementation, and implementation was initiated for 16 during the study period. Of these 16 actions, four required hospital-wide coordination through cross-departmental operational changes or organizational policy revision. The ED recorded 29,023 visits and 184 incident reports during the pre-intervention period and 28,313 visits and 430 incident reports during the post-intervention period. The reporting rate increased from 6.34 to 15.19 per 1,000 visits. In interrupted time series analysis, the post-intervention period was associated with an immediate increase in the reporting rate (incidence rate ratio 2.07, 95 percent confidence interval 1.37 to 3.13). There was no statistically significant change in slope. The overall distribution of the degree-of-harm categories did not differ between periods, and no incident reports were classified in the death category.

Integrating ED peer review and recurring joint incident meetings into a centrally managed incident reporting system provided a structured pathway through which system improvement actions progressed from proposal to approval and initiation of implementation, including actions requiring coordination beyond the ED. Longer follow-up and multicenter studies are needed to assess the sustainability of this approach and its effects on patient-centered safety outcomes.