VitaSignal Evidence in Context

VitaSignal Evidence in Context is an interactive synthetic demonstration for clinical-AI evaluation leads, vendor teams, and health-system reviewers. Its fictional tool, Harbor Draft 1.0, prepares outpatient documentation drafts for staff review and does not write to the chart automatically.

The fictional health system is considering this tool for the first time. The journey starts with no recorded decision. Qualification here means reviewing whether the evidence supports a defined next step, not proving that the tool works or certifying it for patient use.

The synthetic example uses three technical app roles. Clinical AI Buyer Readiness, the buyer-readiness product, is intended to help a clinical-AI vendor organize its intended use, claims, supporting evidence, and unresolved gaps. EHR and Data Pathway is intended to help a healthcare organization examine the local data and workflow dependencies of that use. Governance is intended to help accountable institutional reviewers document the next step, its conditions, responsibilities and unresolved questions.

Clinical AI Buyer Readiness is in private preview on its own website; Governance and EHR and Data Pathway are private prototypes. Every stage here is simulated, not product access.

An internally developed tool can enter institutional review through its project owner rather than a vendor portal. Not every proposed use requires an EHR connection; the review follows the dependencies of the actual use case.

The three are designed to connect through an evidence handoff. This demonstration uses simulated handoffs, not live connections between VitaSignal's products. A vendor-prepared record does not automatically establish local suitability or authorize clinical use.

The vendor packet contains an invented product statement and six fictional outpatient encounters. It contains no inpatient cases, reference-standard scoring, error inventory, or performance measurements. The referenced retention statement was not supplied and remains unresolved.

The visitor can inspect source passages instead of treating a polished claim as evidence. No real patient records are included, and no customer outcome is shown.

An invented local worksheet describes outpatient notes available after the visit, a note-type mapping, staff review, and no automatic chart write-back. These are example checks, not an institutional verification or proof of semantic equivalence.

There is no live EHR connection. The worksheet does not establish inpatient note types, data meaning, required timing, or reviewer capacity.

After reviewing the fictional sources and local workflow, the visitor explicitly acknowledges the conditions and unresolved questions, then records V1 for Morgan Lee, a fictional accountable reviewer. The demo does not record an approval automatically.

V1 says: Proceed with synthetic outpatient evaluation planning only. It does not run the tool, authorize actual testing, permit patient use, allow chart writing or authorize deployment. The missing retention statement, reference-standard scoring, error inventory and performance measurements remain unresolved.

This completes the first-time journey. A real health system may decide to Proceed, Modify, Pause, or Stop based on its evidence and authority. This fixed synthetic case illustrates a planning-only next step; recording a decision does not make a product ready.

After recording V1, the visitor may change the intended setting from outpatient to inpatient. This is an optional reassessment, not a software-platform migration or a requirement to finish the first-time review. The tool identity and supplied source history stay the same, but outpatient evidence does not become evidence of inpatient suitability.

The comparison separates what remains applicable as history, what requires reassessment, what is unknown, and what is still missing. It does not guess missing inputs, label the tool unsafe, or calculate a new clinical performance score.

The recorded V1 decision is limited to fictional synthetic outpatient evaluation planning only. It does not permit actual tool testing, patient use, chart writing, or deployment.

Trying to continue with V1 after selecting inpatient is blocked by scope. V1 remains unchanged in the history; the demonstration does not silently extend its authority to a setting that was never reviewed.

A fictional accountable reviewer can record V2 for the inpatient setting: Pause implementation; prepare an inpatient evidence plan only. This is a simulated decision, not an approval by a real health system.

Recording V2 preserves V1 and carries unresolved findings forward, including the absent retention statement and the inpatient evidence, data-pathway, and reviewer questions. Recording a decision does not supply missing evidence.

This experience illustrates intended technical responsibilities and the scope of a versioned Governance Decision Package. The EHR app supports institutional work; these stages are not a required purchase sequence. It is not a completed live integration, a clinical tool, an assessment of a real vendor, or evidence of real-world performance.

No clinical utility, patient benefit, security certification, regulatory clearance, or production readiness is established. Real institutional review and a human-owned Proceed, Modify, Pause, or Stop decision require the applicable evidence and authorization.