To develop a graded classification system for risks in emergency medicine. To test the inter-user reliability of this classification system.Design:
Prospective collection of data involving emergency department (ED) critical incidents. Derivation of classification system using the collected critical incidents. Comparison of results of classification of a sample of the critical incidents between different users of the system.Setting:
EDs in two teaching hospitals and two district general hospitals (DGHs) in the north west of England.Setting:
Interventions: Observational study.Setting:
Main outcomes: Classification system itself. Results of classification of same critical incidents by different users.Results:
816 critical incidents were identified and used to derive a typology. This typology was found to have inter-user reliability score of 86% (95% confidence intervals 76.4% to 95.6%).Conclusions:
The typology that has been derived is a reliable tool for the classification of risks in emergency medicine.