Skip to content

Defence assurance test catalogue

This is the generated human-readable view of the canonical assurance-test catalogue. Its 13 scenarios are designs, not executed results. A listed test does not demonstrate system conformance, clinical safety or operational acceptance.

The scenarios use synthetic records and representative workflows. Execution still requires an agreed system boundary, clinical-risk controls, test-data governance, measurable thresholds, named evidence owners and authorised pass/fail decisions.

Test-status definition

Status Meaning
designed_not_executed The scenario has been designed, but no evidence of system conformance, clinical safety or operational acceptance has been produced.

Catalogue

AST-001 — Reconcile identity, physical support location, care responsibility, registration and source coverage for a mobile cohort

Status: designed_not_executed
Candidate requirements: REQ-001

Scenario objective

Reconcile identity, physical support location, care responsibility, registration and source coverage for a mobile cohort

Method

Seed synthetic posted, attached, embarked, deployed, temporary, evacuated, duplicate and wrongly registered records; exercise conflict and correction paths

Evidence required for a pass

Each attribute and source is independently visible; conflicts and missing periods cannot appear silently complete; an accountable correction is recorded with provenance

AST-002 — Demonstrate complete safety-search and recall populations despite location and registration variation

Status: designed_not_executed
Candidate requirements: REQ-002, REQ-001

Scenario objective

Demonstrate complete safety-search and recall populations despite location and registration variation

Method

Run screening and readiness queries over a seeded cohort containing eligible, excluded, non-responder, temporary, stale and wrongly registered people

Evidence required for a pass

Displayed denominator, exclusions, source coverage and freshness reconcile to the seed; every unresolved person has an accountable follow-up state

AST-003 — Retrieve and reconcile recent care and the minimum safety dataset across supported record sources

Status: designed_not_executed
Candidate requirements: REQ-003, REQ-001

Scenario objective

Retrieve and reconcile recent care and the minimum safety dataset across supported record sources

Method

Transfer synthetic allergies, medicines, risks, care plans and recent treatment between permanent, temporary, NHS, Defence and receiving services, including one unavailable source

Evidence required for a pass

The receiving view preserves provenance, identifies the latest authoritative values, discloses the unavailable source and time period, and prevents an incomplete record appearing complete

AST-004 — Close a time-critical result safely during workload, recipient absence and delayed connectivity

Status: designed_not_executed
Candidate requirements: REQ-004, REQ-009

Scenario objective

Close a time-critical result safely during workload, recipient absence and delayed connectivity

Method

Inject graded abnormal laboratory and imaging results during routine, night, surge, wrong-recipient and disconnected scenarios; observe routing, acknowledgement, escalation and closure

Evidence required for a pass

Severity and required action remain salient; an accountable owner acknowledges within the defined time or escalation occurs; clinical action and closure are evidenced without equating viewing or filing with completion

AST-005 — Route safety-relevant patient messages, referrals and recommendations through every supported channel

Status: designed_not_executed
Candidate requirements: REQ-005, REQ-009

Scenario objective

Route safety-relevant patient messages, referrals and recommendations through every supported channel

Method

Submit ambiguous and high-risk synthetic content to administrative and clinical entry points; test rejection, reassignment, non-response and record incorporation

Evidence required for a pass

Users can see destination and responsibility; clinical risk reaches accountable triage; rejection or delay escalates; the record preserves content, decision, action and completion

AST-006 — Deliver and reconcile draft, corrected and final transfer records into the receiving workflow

Status: designed_not_executed
Candidate requirements: REQ-006, REQ-003

Scenario objective

Deliver and reconcile draft, corrected and final transfer records into the receiving workflow

Method

Exercise ambulance, aeromedical, civilian and unit-transfer handovers with late finalisation, correction, duplicate delivery and one unavailable interface

Evidence required for a pass

Receiving staff can distinguish draft from final, see completeness and arrival, acknowledge review, reconcile verbal differences and retain the final record longitudinally

AST-007 — Maintain safe hybrid care and time-bounded medication review during digital unavailability

Status: designed_not_executed
Candidate requirements: REQ-007, REQ-012

Scenario objective

Maintain safe hybrid care and time-bounded medication review during digital unavailability

Method

Record a safety-critical recommendation on paper during outage, suspend a medicine, restore service and reconcile competing updates

Evidence required for a pass

Paper presence and authority are signposted; the suspension has an owner and deadline; all content is attributable; restoration produces one reconciled history without silent loss or duplication

AST-008 — Preserve amendment provenance consistently across live, routine, printed, exported and investigation views

Status: designed_not_executed
Candidate requirements: REQ-008

Scenario objective

Preserve amendment provenance consistently across live, routine, printed, exported and investigation views

Method

Create, correct and late-enter controlled records, then compare every supported representation and audit view

Evidence required for a pass

Every view clearly distinguishes original and amended content, author, event time, entry time and amendment status; no export implies that later text existed at the original time

AST-009 — Evaluate whether high-severity decision support is usable and proportionate under representative workload

Status: designed_not_executed
Candidate requirements: REQ-009, REQ-004, REQ-005

Scenario objective

Evaluate whether high-severity decision support is usable and proportionate under representative workload

Method

Scenario-based usability test with varied severity, repeated low-value alerts, override, absent owner, fatigue and assistive-technology use; measure sensitivity, specificity and downstream action

Evidence required for a pass

High-severity hazards are detected and acted on within target; low-value burden and unsafe override remain within agreed limits; every alert has an actionable owned outcome

AST-010 — Prove that existing safety barriers survive CORTISONE migration or have an accepted residual risk

Status: designed_not_executed
Candidate requirements: REQ-010

Scenario objective

Prove that existing safety barriers survive CORTISONE migration or have an accepted residual risk

Method

Build a pre/post-migration control inventory and execute the same clinical-safety regression cases against source, transitional and target states

Evidence required for a pass

Every barrier maps to equivalent or stronger tested behaviour; differences have named owners, documented clinical-safety assessment and authorised residual-risk acceptance before release

AST-011 — Recover safely from partial failure across independently administered transaction endpoints

Status: designed_not_executed
Candidate requirements: REQ-011

Scenario objective

Recover safely from partial failure across independently administered transaction endpoints

Method

Simulate cancel, reissue, dispense, administer and acknowledgement races with retries, duplicate messages, out-of-order delivery and a disconnected endpoint

Evidence required for a pass

Endpoint states and provenance are observable; replay is idempotent; conflicting action is prevented or contained; unresolved divergence is assigned and reconciled

AST-012 — Restore service after prolonged disconnection without silent data loss, unsafe replay or version conflict

Status: designed_not_executed
Candidate requirements: REQ-012, REQ-011

Scenario objective

Restore service after prolonged disconnection without silent data loss, unsafe replay or version conflict

Method

Run offline care with clock skew, duplicate edits, queued tasks and staged reconnection across multiple endpoints

Evidence required for a pass

Authority and fallback are clear during outage; queued work stays visible; synchronisation detects conflicts, avoids duplication and produces an attributable reconciliation record

AST-013 — Share the minimum necessary safety signal across clinical and operational services lawfully and accountably

Status: designed_not_executed
Candidate requirements: REQ-013

Scenario objective

Share the minimum necessary safety signal across clinical and operational services lawfully and accountably

Method

Exercise detention, evacuation and force-protection scenarios with permitted, prohibited, stale and corrected risk signals and role changes

Evidence required for a pass

Only agreed signals reach authorised roles for the stated purpose; provenance and freshness are visible; access and decisions are audited; an accountable reviewer resolves stale or conflicting information

Evidence expected from execution

An executed test record should identify the configuration and interfaces tested, synthetic-data seed, scenario version, participants and roles, observed event trail, human-factors findings, deviations, clinical-risk decision, evidence owner, date and approval status. A demonstration or screenshot alone is not a pass.

Tests may be split into lower-level cases during programme assurance, but their AST- identifier should remain the parent trace to the candidate requirement.

Maintenance

Edit assurance_tests.csv, then run node scripts/generate-defence-pages.mjs. Do not edit this page directly. Preserve retired identifiers rather than reusing them.