Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for medical records clerks

Duplicate Patient Record Agent

Duplicate charts are found within a day and merged only with clear evidence

Duplicate Patient Record Agent: what goes in, what the agent does and what you get

What it does

A patient registered twice ends up with two charts, and results can land in either one, creating care and billing errors. Each day this agent compares new registrations with the master patient index. It scores likely pairs on name, birth date, address, phone and insurance number, allowing for typos and nicknames. If a pair lands in the uncertain band, it pulls more fields such as visit history and guarantor and scores again. Twins share a birth date, address and phone, so a different first name with the same birth date is never proposed without a manual check. It drafts a merge request showing both records side by side with the evidence. The clerk approves every merge, since merging the wrong people is hard to undo.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Daily registrations close 2 USES A TOOL Compare new registrations with the patient index 3 DOES Score each candidate pair on several fields 4 CHECKS THE RESULT Is the score clearly above or below the mergethreshold? If not: pull visit history, guarantor and insurance dataand score again. Back to step 3. 5 CHECKS THE RESULT Could the pair be twins or family members? If not: send the pair for manual check instead of amerge request. Back to step 3. 6 DOES Draft merge request with side-by-side evidence 7 YOU APPROVE Clerk approves or rejects each merge 8 USES A TOOL Log merges and rejected pairs 9 RESULT Merge log and list of rejected pairs
Read the steps as a list
  1. Daily registrations close
  2. Compare new registrations with the patient index
  3. Score each candidate pair on several fields
  4. Is the score clearly above or below the merge threshold?If not: pull visit history, guarantor and insurance data and score again. Back to step 3.
  5. Could the pair be twins or family members?If not: send the pair for manual check instead of a merge request. Back to step 3.
  6. Draft merge request with side-by-side evidence
  7. Clerk approves or rejects each mergeThe agent waits here for your OK.
  8. Log merges and rejected pairs
  9. Merge log and list of rejected pairs

How it decides

It proposes a merge only when the match score is above the set threshold and no field points to a different person, such as a different first name with the same birth date.

  • Never propose a merge when first names differ and birth dates match
  • Uncertain scores get more data before a decision
  • Rejected pairs are remembered so they are not proposed again

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Merge threshold score (default 90)
  • Fields used for matching
  • Uncertain band range
  • Daily or twice-daily scan

What keeps you in control

It always asks you first

  • Merging two patient records

Hard limits

  • Never merges records itself
  • Never deletes either record

It stops when

  • Done: all candidate pairs merged or rejected
  • Stop: a pair has clinical results in both charts on the same day; send to the records supervisor

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA registration for Jon Smyth, born 4 March 1961, scored 78 against John Smith with the same birth date and phone. The threshold check failed because 78 sat in the uncertain band. The agent pulled insurance data, found the same member number and rescored at 96. It drafted a side-by-side request. The clerk approved the merge, and the agent logged it.

More agents for medical records clerks