Modeling AI Triage Impact on Patient Intake
Quantified the impact of AI triage tools on clinician workload and authority delegation.
The Challenge
A regional hospital introduced an AI triage tool into their emergency department intake workflow but lacked data on how it changed clinician decision-making, workload distribution, and the effective authority structure at each step.
The Approach
H-Synth modeled the pre-AI and post-AI intake workflows as documented and practiced variants. NASA-TLX and Trust in Automation instruments were deployed via live observer experiments during shifts. Monte Carlo simulation predicted bottleneck impact under varying patient volumes.
The Results
The hospital discovered that AI triage reduced cognitive workload at initial assessment by 40% but shifted authority delegation patterns in ways that conflicted with their documented protocol — leading to a targeted policy update.