Skilled nursing product in development

Ease currently serves behavioral health organizations and is planning a skilled nursing product. Discuss planned scope and availability.

Health information manager and IT lead

Duplicate Resident Records in a SNF EHR: A Merge and Correction Review

Evaluate resident matching and record correction using duplicate, overlay and cross-facility cases, with a review queue and downstream verification steps.

Quick answer

First determine whether a suspected identity issue is two records for one resident, information entered in the wrong person’s record, or identifiers that legitimately differ across facilities. Authorized staff should review the evidence before any merge or correction. Preserve source identifiers, decision history and downstream relationships, then verify affected clinical and financial workflows. A matching score or similar name is not sufficient evidence to combine records.

Distinguish the problem before choosing the action

Two charts can belong to one person. One chart can also contain another person's information. Those situations call for different investigations. Treating every identity problem as a duplicate can combine records that should remain separate.

The SAFER Patient Identification guide distinguishes duplicates, overlays and cross-facility identifier issues, and addresses their prevention and correction. This article provides an original procurement and operating exercise. Actual record correction should follow the facility's authorized process, with clinical review where care may have been affected.

Begin by identifying the source system, facility and identifier for each candidate. A local record number is not necessarily unique across an operator's acquired facilities. Keep that context in the matching process instead of treating equal numbers as equal people.

Create a review queue with a clear disposition

Each candidate needs a reason for review, supporting evidence, responsible reviewer and final disposition. Restrict access to the information needed for authorized resolution. Avoid circulating full resident charts in general support tickets or email threads.

Candidate What it might mean Review question
Same name, different source identifiers Duplicate or two different people What authoritative evidence distinguishes identity?
One resident has two charts after readmission A new episode may have created a second identity Should this be one person with separate episodes?
A document appears in the wrong chart Possible overlay or misfiled document Which information and downstream work are affected?
Two facilities use local number 1042 Legitimate separate local numbering Were facility and source identifiers preserved?
A changed name produces a new match candidate Demographic update or a different person Can the identity history explain the difference?

Use explicit states such as awaiting evidence, confirmed same person, confirmed different people and escalated for correction. A queue entry closed without a disposition loses the reason a future import should behave differently.

Test a fictional same-person case

Construct two fictional charts for the same resident, with an authorized review decision specifying the surviving identity. One chart contains an admission episode and the other contains a later hospital-return document. The expected result is a coherent person-level record that preserves the distinct episode and source histories.

After the approved operation, search using both original identifiers. Inspect whether the source aliases still lead to the intended person, whether record permissions remain correct and whether the hospital-return document retains its original provenance. A tidy resident list does not establish that all relationships survived.

Run a later import containing the retired identifier. The importer should follow the approved crosswalk or flag the conflict through the established process, rather than silently recreating the duplicate. Record the version of the matching decision used for that import.

Test the different-person and wrong-chart cases

Create two fictional residents with similar names and distinct supporting identity evidence. The application may suggest a possible match, but the review workflow must allow an authorized rejection of the candidate and preserve the reason. Check whether the rejected candidate reappears indefinitely or is incorrectly suppressed after relevant evidence changes.

For a wrong-chart document, ask the vendor to demonstrate its supported correction process. Identify what happens to the original location, attribution, audit record and any downstream copies. Do not assume that moving one visible document corrects an order, result, claim or notification generated from it.

If an erroneous merge has already occurred, test the vendor's actual remediation capability using fictional data. Some systems require specialist intervention. Record that dependency and the facility's interim process; do not advertise an unmerge function simply because a salesperson describes manual support.

Reconcile the affected relationships

A fictional review finds 12 duplicate candidates. Seven are confirmed duplicates, three are different people and two await further evidence. Reporting “12 duplicates resolved” would overstate both the error count and the completed work. The disposition ledger should preserve all three outcomes.

For each confirmed correction, identify the linked episodes, documents, orders, results, billing records and external identifiers that need verification. The appropriate owners should review whether clinical or administrative work occurred while information was misassociated. An administrator should not make clinical conclusions from the identity log alone.

Keep a record of who approved the action, when it occurred, the affected identifiers and the evidence used. Preserve the distinction between the source's original data and the organization's reviewed correction.

Bring identity resolution into platform selection

The migration acceptance guide checks the overall destination and usability of transferred data. This exercise tests the more specific ability to investigate and correct person-to-record relationships, including after normal operations have begun.

Ask each vendor to show the review queue, permission boundaries, supported correction process and downstream checks. Include the health information lead and staff who handle admissions and billing. A matching algorithm's headline accuracy cannot answer whether the facility can safely investigate an ambiguous case and preserve its decision over time.

Questions & answers

Can two resident charts be merged because their names and birth dates match?

Those similarities can identify a candidate for review but may not establish that the records belong to the same person. Use the organization’s approved identity-resolution process and appropriate source evidence before merging.

Is reversing a merge the same as deleting the incorrect record?

No. A correction may need to disentangle documents, orders, results, claims and external identifiers while preserving an audit history. Verify the actual recovery capability and involve the appropriate clinical, records and technical owners.

Primary sources

Use the cited source and its applicable version when making a facility-specific decision. These guides explain operational evaluation methods; they are not individualized clinical, coding, or legal advice.