Chart prep without the clicking

The problem

Medical assistants review the next clinic day's patients for each provider and check which quality measures each patient still needs: screenings, exams, a yearly physical. To do that well they need to know what the EHR already has on file, including the most recent labs. Looking that up means opening each chart by hand, one patient at a time, before the day starts.

What was built

A chart review board fills itself from the EHR schedule and shows, per patient, which measures still apply, with eligibility rules by age and insurance. A physical the patient already attended this year closes the yearly physical measure on its own.

Then an RPA bot was added. It opens each upcoming patient in the EHR, reads their history notes and most recent labs, and matches what it finds to the measures. Medical assistants see what is already on file next to each measure, plus the date of the last labs, right on the review board.

What was hard

An overlay on the EHR's summary screen blocked normal automated clicks, so the bot triggers the link directly. And matching each patient to the right chart had to be nearly perfect: a review attached to the wrong chart is worse than no review. The bot checks in at the start and end of every run, so a silent failure shows up on the admin page the same morning.

Result

The bot matched 1,405 of 1,409 patients to their EHR chart and reads about 160 patients an hour. A backfill of 54,198 earlier appointments closed the yearly physical for 1,239 patients from visits they had already attended.

What is still not measured: how long assistants used to spend looking this up per chart. The logs don't record the earlier process, so no time saving is claimed.

Where the numbers come from

Build notes and the bot's own run logs from October 2026, held in the practice's work log. Patient details never leave the EHR; the bot records only what measure is covered and when.