Duplicate Patient Resolution

This page is the end-to-end reference for how CareLaunch prevents, detects, and resolves duplicate patient records. The Duplicate Patient Resolution tool gives administrators a way to review flagged pairs and resolve them by merging, dismissing, or removing them from the list.

Access: Platform Admins only.

Preventing Duplicates

CareLaunch reduces duplicates at the point patients enter the system, but the behavior depends on how the patient is created.

Imports auto-merge into existing records

When you import patients in bulk (batch or CSV import), each incoming row is matched against your existing patients in this order:

1. Identifier (such as an MRN or external ID)
2. First name + last name + date of birth
3. First name + last name + email

If a match is found, the import updates the existing patient record instead of creating a new one — it does not create a duplicate. This happens automatically and silently.

What this means for re-imports

If you re-import your patient roster, matching patients are merged into their existing records rather than duplicated. This is expected behavior — an import is safe to run again. New rows that don't match any existing patient are created as new patients.

Manual creation and self-registration do not reliably prevent duplicates

Neither manual patient creation nor portal self-registration performs a duplicate check that stops a duplicate from being created — the two paths behave differently, but in both cases a duplicate can still be saved:

• Manual creation (admin interface): As you enter a patient's name, the form may surface a soft "This patient might already exist" hint that links to the matching record. This hint is advisory only — it appears solely on an exact full-name match, it never blocks you from continuing, and it should not be relied on to catch duplicates.
• Self-registration (patient portal): No duplicate check runs at all. A patient who registers again with the same details creates a new record.

Because neither path is a reliable safeguard, duplicates created this way are caught later by the Duplicates tab. Review the Duplicates tab periodically to clean them up.

How the Duplicate List Is Generated

The duplicate list is generated on demand when you open the Duplicates tab — it is not a background report that runs on a schedule. Each time you open the tab, CareLaunch scans your patient records and returns the pairs that may represent the same person.

By default, two records are flagged as potential duplicates when they share the same first name, last name, and gender. Date of birth and email are shown alongside each pair so you can review them, but they are not part of the default matching criteria.

Organizations can request a customized matching query if the default criteria don't fit their needs — for example, to match on additional fields. Contact support to customize matching for your organization.

The list only includes pairs that have not yet been resolved. Once a pair is merged, marked as not a duplicate, or deleted, it no longer appears in the active list.

Accessing the Tool

Open the Dashboard and select the Duplicates tab. The tool displays all unresolved potential duplicate pairs for your organization.

Use the search bar to filter by patient name if you are looking for a specific record.

Understanding the List

Each row in the list represents two patient records that have been flagged as potential duplicates. The table shows:

Column	Description
Patient 1	Name, date of birth, and contact info for the first record
Patient 2	Name, date of birth, and contact info for the second record
Date	When the duplicate pair was first detected

Rows where both patients share the same date of birth are highlighted to make them easier to spot.

How Pairs Are Scored

Each flagged pair is scored internally based on how closely the two records match. This score is not shown as a column in the list — instead, it drives a warning inside the comparison dialog (see below) when a pair is very likely a true duplicate.

With the default matching criteria (first name, last name, and gender), pairs fall into one of two levels:

Level	Meaning
Exact	The display text for both records is identical
Medium	The records match on the criteria but their display text differs (for example, different capitalization or a middle name)

Higher- and lower-confidence levels only arise for organizations using a customized matching query; with the default matching you will only see Exact and Medium pairs.

Reviewing a Pair

Click any row to open the comparison dialog. This shows a side-by-side view of both records including name, date of birth, gender, phone number, and managing organization. You can click View Profile on either side to open the full patient record in a new tab.

Expand the Technical Details panel to see a field-by-field comparison of both records alongside a preview of what the combined record would look like after a merge. A warning is shown if the similarity score is high, prompting extra care before taking action.

Choose an action from the dropdown at the top of the dialog before confirming:

• Merge — combines both records into one (see below)
• Mark Not Duplicate — dismisses the pair as a false positive (see below)

Click Confirm to apply the action, or Cancel to close without making any changes.

Merging Records

Selecting Merge and confirming combines both patient records into a single record. Here is exactly what happens:

• Patient 1 is the surviving record. All of Patient 2's data — appointments, observations, messages, tasks, payments, and other associated records — is re-linked to Patient 1.
• Patient 2's record is then permanently deleted. Its identifier is kept on the surviving record for traceability, so you can still trace history back to the record that was merged away.
• The duplicate entry is removed from the list after a successful merge.

Merging is permanent

Merging cannot be undone, and the merged-away record is permanently deleted. Review both records carefully in the comparison dialog before confirming.

Very large patient records can occasionally fail to merge and return an error. If a merge fails, contact support rather than retrying repeatedly.

Marking as Not a Duplicate

If the two records belong to different patients, select Mark Not Duplicate and confirm. This dismisses the pair and removes it from the active list. No patient data is changed or deleted — the two records remain separate and independent.

Use this option when the system has flagged two records that look similar but are genuinely different people.

Deleting a Duplicate Entry

Deleting removes the duplicate pair from the list without merging or altering either patient record. This is useful when a flagged pair is outdated, invalid, or was created in error.

To delete a single entry, open the comparison dialog and close it, then use the bulk delete flow described below for individual items by selecting just that one row.

note

Deleting a duplicate entry does not delete the patient records themselves. Both patients remain in the system unchanged.

Bulk Operations

To act on multiple pairs at once, use the checkboxes on the left side of the table to select rows. The toolbar updates to show the number selected and reveals two bulk action buttons:

Bulk Merge / Mark Not Duplicate

1. Select one or more rows using the checkboxes.
2. Click Merge in the toolbar.
3. In the confirmation dialog, choose an action from the dropdown:
   • Merge — merges all selected pairs
   • Mark Not Duplicate — dismisses all selected pairs as false positives
4. Click Confirm to proceed.

Bulk Delete

1. Select one or more rows using the checkboxes.
2. Click Delete in the toolbar.
3. Confirm the deletion in the dialog.

Bulk delete removes the selected duplicate entries from the list without affecting the underlying patient records.

Tips

• Start with Exact matches. Records with an Exact similarity score are the strongest candidates for merging.
• Check date of birth. Highlighted rows (matching date of birth) are more likely to be genuine duplicates.
• Use View Profile to open a patient's full record in a new tab before deciding, especially for Medium or High similarity scores.
• High similarity does not guarantee the same person. Even when many fields match, verify both records before merging — merges cannot be undone.
• Mark rather than delete when unsure. Marking a pair as Not Duplicate keeps a record of the review decision.