A Study of Emergency Department AI Prediction Impact

Completed

Brief Summary: The purpose of this study is to evaluate the impact of an AI admission prediction tool on the number of preventable hospital admissions, emergency department (ED) length of stay, when the predictions are displayed only to a dedicated ED triage team. Also, to evaluate user perceptions of the AI tool among the triage team users and medical officer of the day users. Additionally, to evaluate any impact of the AI tool on the number of interventions performed by the triage team, and to evaluate the impact of the tool on time-to-admission after an admission order is placed.

Recruitment & Eligibility

Inclusion Criteria

For the survey component, any HIM clinician that works a shift in the triage area, ED physicians, and the medical officer of the day will be included.
For length of stay data, adult patients registered in the Mayo Clinic-Rochester St. Mary's Emergency Department will be included.

Exclusion Criteria

For the survey, clinicians not working a triage shift during the study period will be excluded.
For the length of stay analysis, only adult ED patients will be included, who do not triaged to the behavioral health/psychiatry pathway, nor patients who are triaged to the Emergency Department observation pathway.

Primary Outcome Measures

Name	Time	Method
Hospital Admissions	282 days	Number of avoidable admissions prevented as a fraction of all ED patients in a day, specifically, the number of patients who were seen by the SAPPHIRE triage team and discharged home

Secondary Outcome Measures

Name	Time	Method

Receive instant email notifications about changes in this clinical trial

Receive instant email notifications about changes in this clinical trial