Projektname
Reducing Mistriage in Emergency Departments: A Path to Improved Triage Accuracy through Real-time Evaluation and Artificial Intelligence 

Projektverantwortliche
Christian Nickel
Philippe Claude Cattin
Michael Simon
Franziska Zúñiga

INS-Projektteam
Serena Sibilio
Bastiaan Van Grootven
Michael Simon
Franziska Zúñiga
Ursula Feuz

Projektpartner
Universitätsspital Basel (Christian Nickel, Roland Bissinger, Jens Eckstein)
Departement Biomedical Engineering, Universität Basel (Philippe Claude Cattin, Annette Christine Mettler)
Kantonsspital Aarau (Guglielmetti Sonia),
Luzerner Kantonsspital (Michael Christ)
Kantonsspital Baden (Markus Schwendiger)
Universitätsklinik für Notfallmedizin Inselspital Universität Bern (Wolf Hautz)
Kantonsspital Liestal (Nicolas Geigy)

Orte der Datenerhebung
Notfallzentrum Universitätsspital Basel
Notfallzentrum Kantonspital Aarau
Notfallzentrum Luzerner Kantonspital
Notfallzentrum Kantonspital Baden
Notfallzentrum Universitätsspital Bern
Notfallzentrum Kantonsspital Baselland

Laufzeit
2025 bis 2029

Projektbeschreibung
Hintergrund
The demand for emergency care varies and can change rapidly. Triage is the first step in the emergency department (ED) process, aiming to identify patients needing time-critical interventions, relieve suffering, ensure safe discharges, and efficiently use available resources. Despite the recognized importance of accurate triage, about 20% to 30% of patients experience incorrect triage level assignment, known as mis-triage. Factors leading to poor triage performance remain unclear. Current studies on triage evaluation may not accurately reflect real-world practices, and there is no standard to determine patient acuity. The use of surrogate outcome measures such as hospital admission, Intensive Care Unit admission, and mortality rates as markers of (predictive) validity is controversial. Many triage evaluation studies rely on retrospective data, written scenarios, or electronic health records, which may be biased. Furthermore, few studies have assessed the clinical impact of mis-triage, leaving its causes and consequences unclear.

Zielsetzung
The RemEDy project aim to improve triage processes in EDs by defining optimal processes and quantifying confounding factors that impede effective triage. The RemEDy project will achieve this trough the following aims: 1) To develop an evidence- and consensus-based triage evaluation instrument for real-time assessment 2) To identify and analyze clinician’s and patient’s characteristics and ED metrics related to mis-triage and its consequences 3) To develop and validate an AI algorithm to support triage decision-making 4) To develop a targeted training program for triage clinicians and assess its effectiveness.

Methode
Identification of mis-triage and improvement of triage processes in the ED will be addressed with a multi-phased mixed methods project methodology.

1. Scoping Review and Delphi Study: In work package (WP) 1 a scoping review, following Joanna Briggs Institute guidelines, to identify existing knowledge on triage evaluation will be conducted. Atriage evaluation instrument will be developed, guided by the results of the scoping review and the consensus opinions of expert clinicians and patient representatives determined by a modified Delphi study.
2. Multicentric Observational Study: In WP2, trained triage experts will use the developed triage evaluation instrument to identify mis-triage in ‘real-time’ during routine clinical practice in the ED, which has never been done before. This will be a multicentric, prospective observational study, conducted across seven EDs in Switzerland over 6 months.
3. Artificial Intelligence Algorithm Development and Validation: In WP3, AI triage algorithm to recommend triage scores and predict mis-triage, using both pre-existing and real-time data from the observational study will be developed. The algorithm will be validated and updated with new clinical data to enhance predictive performance and clinical decision support, providing recommendations and warnings about potential over- or under-triage.
4. Training Development and Pragmatic Trial: In WP4, a training program for triage clinicians will be created using the Behavioral Change Wheel methodology, followed by a pragmatic trial to assess its effectiveness.

Erwarteter Nutzen / Relevanz
RemEDy aims to improve the quality and safety of care in EDs through enhanced accuracy in triage. A comprehensive understanding of mis-triage and other possible triage system defects will allow for the identification of potential areas for improvement, leading to the development of measures to avoid them.