SNF EHR Demo Script: Follow One Resident Episode and Its Exceptions

Prepare one reusable synthetic packet

Create resident TEST-R17, prior episode TEST-E04, new referral TEST-F21 and planned episode TEST-E09. Include a two-page fictional referral document with page two deliberately missing, a historical document clearly dated to E04, and a corrected demographic value. Add a second submission that is an update to F21, not a second opportunity. Do not put real resident information in a sales environment.

Give every supplier the same packet, role list and starting conditions. Require a clean reset before repeat attempts. In the historical HHSC Oahu Region LTC EHR RFP, 2016, the purchaser specifies demonstration scenarios; the practical lesson is that the buyer can control the case rather than accept a tour of unrelated screens.

Run this ninety-minute agenda

This is an original procurement exercise. Clinicians supply safe example content and assess clinical meaning; it is not an admission or treatment protocol.

The general demonstration establishes continuity across functions. Detailed claim generation, payment posting and financial expected results belong in a separate finance acceptance session. A successful general demo should not be reported as proof that the revenue cycle is ready.

Observe the user's explanation after the action

Use one row per task: input version, role, device, start/end time, result, prompts received, mistaken interpretation and work left for the next person. Ask “What state is this case in now?” before coaching. Reaching a screen while believing an unsent draft was transmitted is a different result from understanding the state correctly.

Suppose three synthetic attempts take 45 seconds unaided, 30 seconds after two presenter prompts, and 70 seconds unaided with a correct recovery from a wrong filter. An average of 48.3 seconds hides the two findings that matter: one user required help and another could recover. Report those observations alongside time. Do not extrapolate three attempts into a productivity claim.

The 2025 SAFER System Management guide includes usability, physical environment and user participation in system testing. Apply that bounded guidance by using the actual small-screen device or shared workstation planned for the facility. Include keyboard navigation, enlarged display and error-message recovery as relevant to participating users. These checks do not constitute a complete formal accessibility assessment.

Spend the interruption on a consequential boundary

An independent facility should have the covering coordinator take over without the primary coordinator explaining the case. This tests whether concentrated local knowledge can be transferred. A group should move the referral from a regional queue to Facility B, using a role that cannot edit Facility A. This tests whether the operational handoff and access boundary agree.

In both cases, retain the same missing page. A presenter who adds it before the handoff has changed the task. Ask the new user to locate the outstanding request, determine where a response should arrive, and describe what they cannot yet conclude. Explicit uncertainty can be the correct output.

Turn the result into a bounded next decision

Classify findings as demonstrated, assisted, failed, not tested or outside quoted scope. A single score compresses too much: “not tested” is neither a pass nor evidence of product failure. Separate a mislabeled control from a missing permission and from disagreement over the intended workflow because each requires different work.

For a failed task, name the smallest repeat that would answer the question. If a revised configuration repairs routing, repeat that task and the next-role retrieval, using the original packet. If the solution adds an external service, update both scope and cost. If no acceptable route exists for an essential workflow, the committee has learned enough to narrow or reject that option.

Save the packet and result sheet with the selection record. They become useful implementation examples because they preserve the buyer's actual questions, including the interruptions that a polished demonstration tends to omit.