Skilled nursing product in development

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

IT lead and clinical evaluation team

SNF EHR Test Data: Build Realistic Scenarios Without Using Resident Charts

Build fictional resident scenarios for SNF software evaluation, with linked events, expected results and safeguards against accidental live messages or claims.

Quick answer

Use wholly fictional resident scenarios for initial SNF EHR demonstrations and acceptance exercises. Define the relationships, dates and exceptions needed to test the workflow, and record the expected result before the demonstration. Keep test accounts and external destinations controlled. Removing a name from a real chart is not enough to establish HIPAA de-identification; any use of actual resident information requires the appropriate privacy and security review.

Model the workflow, not a real person's history

A useful demonstration needs realistic relationships and exceptions. It does not need a resident's actual chart. Construct a fictional episode containing an admission, a changed order, a missing document and a later correction. The evaluation team should know what the system is expected to preserve at each step.

The HHS de-identification guidance describes Expert Determination and Safe Harbor. Replacing a resident's name is not equivalent to completing either method. This guide recommends invented data for early procurement exercises; it does not claim that synthetic labeling makes copied clinical records safe to disclose.

Avoid copying real free text and changing only demographics. Unusual histories, dates, images or narrative details may retain identifying information. Write the scenario from scratch with the clinical lead checking its plausibility. No test scenario should become an instruction for providing care to a real resident.

Give the scenario a defined purpose

Start with a purchasing question. If the concern is missed handoffs, include an unresolved task that crosses shifts. If the concern is payer changes, include a dated change and an expected finance review. More fabricated fields do not necessarily create a better test.

Create a small fixture register:

Fixture element Example of what to define Expected evidence
Resident identity Two fictional people with similar names Records remain distinguishable
Episode One admission followed by a hospital absence Episode and movement history remain traceable
Document Original and corrected versions Staff can identify the current version and prior source
Order New instruction replacing an earlier instruction Authorized workflow shows the applicable version
Unfinished work One item awaiting an outside response Named owner and unresolved state remain visible
Financial event A fictional claim rejection Staff can find the original transaction and correction

Use local policies and qualified reviewers to define clinical expectations. The fixture should test the software's handling of information, not make up treatment instructions or coding rules.

Write an expected-results sheet before the demo

A fictional case contains two residents called Jordan Lee with different dates of birth and separate source identifiers. One returns from a hospital with a corrected document. The other has no hospital return. The expected result is that the return attaches to the correct person and episode, the document's source is visible, and the other record remains unchanged.

The same case includes an unresolved task assigned to a staff member whose shift ends. Test which receiving role can see and act on it. Do not allow the presenter to delete the difficult task and restart with a cleaner case. Record any intervention because it explains what the observed result required.

An expected-results sheet needs the step, role, input, expected state, observed state, evidence location and disposition. A screenshot can show a screen state, but it may not prove that a downstream recipient received the correct message. Decide which evidence is adequate for each step.

Prevent the test from reaching live destinations

Before any execution, verify the environment with the vendor and relevant partners. Confirm that test orders, claims, portal messages, emails and interface events cannot be mistaken for real work. Use partner-approved test configurations where external exchange is necessary; a local “test” label may not be understood by another system.

Keep a destination checklist that includes every outbound channel enabled for the scenario. Identify who can run tests, reset the environment and inspect delivery evidence. If a destination cannot be safely isolated, demonstrate that portion through a documented simulation and mark live integration unverified.

The SAFER System Management guide addresses configuration and testing of EHR components. The fixture and destination controls here are original evaluation practices, not a certification or a guarantee of regulatory compliance.

Make the result reproducible

Version the fixture and expected-results sheet together. Record the product version, configuration, modules and role permissions used. If a vendor fixes an issue, rerun the relevant scenario with the same input and preserve both results. Changing the fixture to remove the failure invalidates the comparison.

Keep test accounts and credentials out of public worksheets. For distributed evaluation teams, designate one owner for resetting cases so an earlier demonstration does not change the starting state of the next one.

A clean final packet contains the fictional scenario, its purpose, the expected results, observed exceptions and remaining dependencies. It lets the committee repeat the resident-episode demonstration consistently and carry useful cases forward into implementation acceptance testing.

Questions & answers

Is a real chart safe for a vendor demo if the resident name is removed?

Not necessarily. Other identifiers, free text, images and combinations of facts can identify a person. Use invented scenarios for early evaluation and follow the appropriate review process for actual resident information.

Can a completely fictional record still cause a real operational problem?

Yes. A test record connected to live interfaces can generate messages, orders, claims or notifications. Confirm the test environment, permitted destinations and external partner setup before executing a scenario.

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.