Information collected on the notification page includes:

• generic information about the incident, (e.g. the date, time, medical specialty, notifier designation, what happened, what was the outcome), and
• specific information related to emergency medicine, (e.g. the phase of care where the incident occurred, the ED role delineation and whether or not the incident involved known ED patient safety problems, such as readmission, absconding, communication failures).

Once collected, incidents are classified by an expert panel appointed by the Australasian College for Emergency Medicine using a pre-defined patient safety classification.

Other information that can be classified includes: contributing factors, severity and outcome of the incident, mitigating circumstances and how the incident could have been prevented. Information on when or where the incident was initiated and detected is also obtained. This is useful to determine how far downstream an error can travel before it is detected, and provides an indication of how resilient a system is. For example, if a patient is incorrectly identified on a booking system at triage, it would be safest for the patient if this error was detected when the patient is first reviewed by medical or nursing staff, rather than at a later stage in the patient care (e.g. before or after imaging or pathology has been performed, or on admission to the hospital or discharge from the ED). The shorter the duration between error initiation and detection, the greater the system resilience and the lower the likelihood that patients will be harmed from the error. Currently, there is little known about system resilience in ED, therefore this data will be particularly important to capture.