Edmonton Zone triage project (EZ Triage)

Background: Canadian Emergency Departments (ED) use the five-point Triage Acuity Scale (CTAS) to prioritize patients according acuity. CTAS scores are used make decisions on patient flow, staffing complement, and funding. Variations in triage can lead mis-categorization delayed care for critically ill patients. Edmonton zone quality improvement audits reduced high low-risk vital sign errors (86% 78%, respectively), increased between-nurse consistency, levelled burden individual nurses. However, previous were time-consuming required auditors with clinical, data management, analysis expertise, preventing their wide-scale adoption.
 
 Implementation: This project utilizes AHS AIW implementation framework. It is extension of a audit feedback tool that was developed validated over three-year window (Cotton et al., 2021). current scaling-up QI framework Tableau dashboard will allow ED administrators educators efficiently examine nurse-level department variation craft local measures improve accuracy.
 The newly uses raw EPIC retrieved from warehouses. translates python script duplicates existing Excel Boolean logic generates binary output variables. These variables then loaded into preformatted displays both variances (trimmed mean standard deviation) following variables: low acuity error rates, overrides, sepsis alerts. (and visualizations) users filter by nurse type. They include all relevant triage-associated be near real-time monitoring or downloaded additional department-level analyses.
 Patient Family Engagement: has consistently been identified as an area concern hospital administrators, clinicians, Most complaints received hospitals about generated at triage. Literature suggested cognitively demanding nurses, have experienced bias during benefit minimizing errors, ensuring consistent triage, allowing other sites across province engage similar efforts. Because these efforts require review patient-sensitive data, not included analyses. Evaluation Methods: assessed two ways: soliciting clinical end-users ease operability dashboard, assessing reliability comparing it definitions screening.
 user interface system refined system. Each refinement cycle accompanied further feedback. A random 10,000-patient sample perform assess performance tool. calculated sensitivity, specificity, accuracy, agreement expert raters. Between-group comparisons automated systems gold-standard manual reviews performed using Cohen’s Kappa.
 Results: Feedback collected emergency departments. Refinement cycles resulted allows decision-makers compare practice nursing level while offering near-real-time access visit-level data. addition Systemic Inflammatory Response Syndrome (SIRS) related model. There 10,000 visits considered analysis. Patients less than 18 years age (3%, n=306 visits) incomplete (n=86) removed. final 9608 visits.
 Four categories assessed: High-acuity, low-acuity, pain, SIRS-criteria. Automated screening had 100% accuracy (κ =1, p<0.001) outperformed high-risk (sensitive=97.8%, specificity=100%, accuracy=99.98%; κ =0.989, p,0.001) (sensitivity=98.9%, =0.994, compared review. Computer (sensitivity=71.3%, specificity=96.7%, accuracy=96.27%; =0.39, assessment pain (sensitivity=52.0%, specificity=97.4%, accuracy=96.58%; =0.335, p<0.001). Expert (sensitivity=100%, specificity=99.9%, accuracy=99.9%; =0.962, p<0.001); computer (sensitivity=7.9%, accuracy=98.91%; =0.145, SIRS-related errors. Advice Lessons Learned:
 Maintaining integrity multiple rounds clinician input.
 small subsequent exclusion pediatric detections thresholds levels modifiers need validation before being patients.
 Further refinements needed incorporate previously text-parsing algorithms detection pain-related errors.