New Patient Intake

Preventing Duplicate Patient Records

Stop duplicate medical records at the source: search rules before creating a chart, field standards, a merge checklist and a monthly cleanup routine.

AssistCall research teamPublished Reviewed by Faruque Amin, VP of Sales and Marketing6 min read

On this page
  1. How duplicates get made
  2. Which fields disagree between duplicates
  3. Search rules before creating a chart
  4. The phone call is where most duplicates start
  5. Field standards that make matches easier
  6. How to merge safely
  7. A monthly cleanup routine
  8. Where duplicates show up later
  9. Frequently asked questions
  10. References
New Patient Intake: Preventing Duplicate Patient RecordsArticle coverNEW PATIENT INTAKEPreventing DuplicatePatient RecordsAva Medical IntakeInsightsNew Patient Intake: Preventing Duplicate Patient RecordsArticle coverNEW PATIENT INTAKEPreventing DuplicatePatient RecordsAva Medical Intake

Key takeaways

  • Most duplicates are created at registration by a search that missed an existing chart, not by the software.
  • Search on date of birth plus a partial last name, never on full name alone, before creating any new record.
  • Middle names and Social Security numbers were the fields most often inconsistent between duplicate pairs in published research.
  • Merging is a clinical safety task. Use a checklist, two people and a log, and never merge on a hunch.

How duplicates get made

A duplicate chart is born in about four seconds. A patient calls, the scheduler types the name as heard, the search comes back empty and a new record is created. The old chart, with the allergy list, the last colonoscopy and the unpaid balance, sits one letter away. The new chart starts collecting results, messages and claims of its own.

Published estimates put the problem in the high single digits. Researchers studying master patient index data cite an average duplicate record rate of about 8 percent for US health care organizations[1]. Small practices often assume they are immune, but every office has the same ingredients: a busy phone line, common surnames, nicknames and patients who have been away for a few years.

The flow of new records keeps the risk constant. New patients made up 16.8% of US physician office visits in national survey data[2], and returning patients who were inactive long enough can look new to a rushed search.

Which fields disagree between duplicates

The same research looked at nearly 400,000 confirmed duplicate pairs and measured how often each identifying field differed between the two records[1]. The result points straight at the front desk habits worth changing.

How often each identifying field differed between confirmed duplicate recordsShare of nearly 400,000 confirmed duplicate pairs with a discrepancy in each field
How often each identifying field differed between confirmed duplicate recordsBar chart. Middle name: 58.1%; Social Security number: 53.4%; Last name: 23.7%; First name: 22.6%; Date of birth: 6.2%.Middle nameMiddle name: 58.1%58.1%Social Security numberSocial Security number: 53.4%53.4%Last nameLast name: 23.7%23.7%First nameFirst name: 22.6%22.6%Date of birthDate of birth: 6.2%6.2%How often each identifying field differed between confirmed duplicate recordsBar chart. Middle name: 58.1%; Social Security number: 53.4%; Last name: 23.7%; First name: 22.6%; Date of birth: 6.2%.Middle nameMiddle name: 58.1%58.1%Social Security numberSocial Security number: 53.4%53.4%Last nameLast name: 23.7%23.7%First nameFirst name: 22.6%22.6%Date of birthDate of birth: 6.2%6.2%

Source: American Health Information Management Association (via PubMed Central), 2016[1]

Show the data
ItemValue
Middle name58.1%
Social Security number53.4%
Last name23.7%
First name22.6%
Date of birth6.2%

Middle names and Social Security numbers disagreed most, because they are often skipped, abbreviated or entered later. First and last names disagreed in more than a fifth of pairs, which reflects spelling by ear, nicknames, hyphenated surnames and name changes. Date of birth was the most stable field, which is exactly why it is the best search anchor.

Language plays a part too. More than one in five people aged 5 and older in the US speak a language other than English at home[3], and names transliterated from other scripts are often spelled several ways by different staff members. Without a rule to follow the ID's spelling, each spelling becomes a new chart.

Search rules before creating a chart

  1. Search on date of birth first

    Enter the date of birth plus the first two or three letters of the last name. This catches most spelling variations without returning hundreds of results.

  2. Search again on phone number

    If nothing matches, search the mobile number. Patients keep phone numbers longer than addresses and often longer than surnames.

  3. Try the first-name swap

    Search with first and last names reversed, and with common nicknames, before concluding the patient is new.

  4. Ask the magic question

    “Have you ever been seen here before, even years ago, or under a different last name?” Many duplicates are prevented by this one sentence.

  5. Create only when every search is empty

    Record which searches were run in the new record's registration note, so a later reviewer can see the check happened.

Make one person responsible for creating new charts during peak hours if your volume allows. Fewer hands creating records means fewer duplicates.

The phone call is where most duplicates start

Schedulers on the phone have the least information and the most pressure. They cannot see an ID, the caller may be a family member, and the next line is already ringing. That combination is why many practices find that new duplicates cluster around phone booking rather than walk-in registration.

A short phone script removes most of the guesswork:

  1. “Can I have the patient's date of birth first?” Starting with the date narrows the search before a name is spelled.
  2. “And the last name, spelled out for me?” Ask for spelling every time, even for names that sound simple.
  3. “What is the best mobile number for the patient?” Search on it before typing anything else.
  4. “Has the patient been seen here before, or at any of our other locations, maybe under another name?”
  5. “I am creating a new record, so I will read back what I have.” Read back name, date of birth and phone in full.

When the caller is not the patient, record the caller's name and relationship separately from the patient's details. A common error is registering the parent's date of birth or phone as the child's, which creates a record that will never match the child's insurance card. If the caller is unsure of any detail, mark the record as provisional and confirm it at arrival instead of guessing.

Track duplicates by channel for a month: phone, walk-in, online booking and referral. The channel with the highest count is where coaching and script changes will pay off first.

Field standards that make matches easier

  • Legal name exactly as the government ID shows it, with the preferred name in its own field.
  • Middle name in full when the ID has it, never a guessed initial.
  • Suffix in the suffix field, so a father and son at the same address stay distinct.
  • Previous names recorded as aliases rather than overwritten.
  • Mobile number in the mobile field, landline in the home field, so search finds them.
  • Address formatted to the postal standard, including apartment numbers in the second address line.

Nearly all office-based physicians now use an EHR[4], and most of those systems include a duplicate warning when a new record looks similar to an old one. The warning only works if the fields are consistent. Teach staff to stop and read the warning instead of clicking through it.

How to merge safely

Merging two charts touches medications, allergies, problem lists, results and billing history at once. A wrong merge, where two different people become one record, is worse than the duplicate it was meant to fix. Treat it as a controlled task with a checklist.

A two-person merge checklist
StepWhoWhat to confirm
Identify candidate pairFront desk or data quality leadDate of birth, name variants, phone, address, insurance member ID
Clinical reviewNurse or medical assistant leadAllergies, medications and problems are consistent with one person
Pick the surviving recordData quality leadThe older record with the most history usually survives
Merge in the systemTrained user with merge rightsUse the EHR's merge tool; never retype data by hand
Notify billingData quality leadOpen claims and balances move to the surviving account
Log the mergeData quality leadBoth record numbers, date, reason and who approved it

If the clinical review finds anything that does not fit, such as different blood types, conflicting surgical histories or a different listed sex at birth with no explanation, stop. That pattern suggests two people or a stolen identity, not a duplicate. Escalate to the practice manager.

A monthly cleanup routine

Prevention never catches everything. A short, regular cleanup keeps the backlog small enough that nobody dreads it.

  1. Run the EHR's potential duplicate report, or a saved search for records sharing a date of birth and phone number.
  2. Sort by most recent activity, so records with upcoming appointments get reviewed first.
  3. Work through the list with the merge checklist, a fixed number of pairs per session.
  4. Mark pairs confirmed as different people so they stop appearing on the report.
  5. Record the count of new duplicates found this month and who created them, then coach privately.

Example: a fictional three-provider family practice ran its first duplicate report and found a long backlog of likely pairs, most created by the phone line during flu season. It set a rule that only two trained schedulers could create new charts between 8 and 10 AM, added the “ever been seen here before” question to the script, and worked the backlog in half-hour sessions every Friday. Within a quarter, new duplicates each month had fallen to a handful.

Where duplicates show up later

Duplicates rarely announce themselves. They show up as a patient who says the doctor never got their results, a lab result with no matching order, a statement sent for a balance the patient already paid on their other account, or two recall reminders for the same mammogram. Ask staff to report each of these as a possible duplicate rather than just fixing the symptom.

Duplicate prevention sits inside the wider intake process. The guides on patient identity verification and scanning ID and insurance cards cover the checks that keep the first record right, and registration errors and claim denials shows what happens to billing when they fail.

Frequently asked questions

What is an acceptable duplicate record rate?

Aim as low as you can measure. Published research cites an average of about 8 percent for US health care organizations, and many quality programs set targets well below that.

Who should be allowed to merge charts?

A small number of trained users, with a clinical reviewer checking medications, allergies and problems before each merge. Front desk staff can flag pairs but should not merge on their own.

Can we delete a duplicate chart instead of merging?

Generally no. Results, messages and claims may be attached to it. Use the EHR's merge function so the history is preserved and an audit trail exists.

Does searching by Social Security number prevent duplicates?

It helps when the number is present, but it is often missing or mistyped. Date of birth plus partial last name and phone number is more reliable as a first search.

References

  1. American Health Information Management Association (via PubMed Central) (2016). Why Patient Matching Is a Challenge: Research on Master Patient Index (MPI) Data Discrepancies in Key Identifying Fields (Perspectives in Health Information Management) (opens in a new tab) ↑

  2. CDC National Center for Health Statistics (2019). National Ambulatory Medical Care Survey: 2019 National Summary Tables (opens in a new tab) ↑

  3. U.S. Census Bureau (2017-2021). New Data on Detailed Languages Spoken at Home and the Ability to Speak English (press release CB25-TPS.40) (opens in a new tab) ↑

  4. Office of the National Coordinator for Health IT (ASTP/ONC), HHS (2024). Office-based Physician Electronic Health Record Adoption, 2008-2024 (ONC Data Brief) (opens in a new tab) ↑

Book a demo

See Ava handle your calls.

Book a short demo. We’ll walk through your calls and workflow, and agree what Ava should take care of.

Book a demo