Patient Summary
AI LLM
Client
SiteRx
Role
Lead UX Designer
Duration
10 months
Challenge
Our clinical excellence staff spent significant time manually reviewing lengthy, fragmented patient records scattered across multiple notes, results, and visit histories in order to assess a case. This slowed down review turnaround and creates risk of overlooking relevant clinical context buried in the record. We needed a way to quickly surface a concise, accurate summary of each patient's history to support faster, more confident case review.
Approach
Conducted 4 user interviews to better understand their current workflow and the pain points.
Mapped out users' workflow to identify where in the process they experienced the most friction based on user interviews and observations.
Worked alongside product, data scientists, and engineers to iterate on functional and compliant designs to fit the needs of users.
Ran an A/B test with users to figure out which of the 2 design directions users felt was the most intuitive via prototypes.
Solution
Developed and designed an internal LLM-powered review tool that summarizes each patient's record — condensing notes, results, and visit history into a concise, structured summary. The tool surfaces relevant clinical context and flags items that may need attention, allowing clinical excellence staff to review cases faster and with greater confidence, while still verifying findings against the source record.
Results
60% reduction in task completion time within 2 quarters.
Increased reviews per week by around 20%.
Built and fostered user trust and adoption for using an AI tool.