Skip to content
Music-Centred Dementia Care
Menu

Research · act-api · draft

ACT-API — integration

Nearing completion. Taking ACT outputs to the point of care.

Summary

ACT-API is the integration layer exposing ACT outputs — alerts, trend summaries, domain trajectories — to care-record systems under a licence model, so anticipatory insight arrives where care is delivered rather than in a separate dashboard. Nearing completion in 2026.

What ACT-API is

ACT-API is the integration layer of the MCDC programme. Its purpose is specific: to expose outputs from the ACT analytics platform — alerts, trend summaries, domain-specific trajectories, ACT summary reports — to the care-record systems that practitioners actually use at the point of care. A good observation instrument and a good algorithm lose most of their value if their outputs live in a separate dashboard nobody has time to check. ACT-API is the piece that makes the dashboard unnecessary: the insight appears where the care record already is.

It is deliberately a layer, not a product. It does not add a new interface; it adapts ACT's structured outputs for existing clinical / care-record systems via a licence-model API.

Current status

Beta — April 2026. The underlying ACT platform and observational app are live (see ACT); ACT-API is the remaining piece in the Component 3 stack and entered beta in April 2026. The priority integration target post-pilot is Person Centred Software (PCS), the most widely used care-management platform in Welsh care homes. Further named consumer integrations will be confirmed as the API stabilises through beta.

[PLACEHOLDER: confirmed first-production-integration date]

Intended design

Three principles shape the design, drawn from the white paper's Section 6:

  1. Standards-compliant by default. The ACT data dictionary has been designed with HL7 FHIR compatibility in mind as the target API standard, SNOMED CT as the clinical-terminology vocabulary, and openEHR archetypes for structured clinical data. These choices are aspirational during the Pendine Park pilot and will be implemented as the API moves into production.
  2. Care-record-first integration. The priority integration target post-pilot is Person Centred Software (PCS), the most widely used care-management platform in Welsh care homes. Other consumer integrations follow on a case-by-case basis, with a common output contract regardless of the target system.
  3. Human judgement primacy. ACT-API does not deliver automated clinical decisions. Outputs are explainable, traceable to the specific data points that produced them, and surfaced for clinical validation before action. This is a design rule, not a technical constraint.

What ACT-API changes at the point of care

Where today an alert lives in an ACT dashboard a care-home manager may or may not open, tomorrow — with ACT-API integrated — the alert appears alongside a resident's other care-record entries. The integration changes the workflow, not the judgement. The clinical lead still decides what the alert means for this person on this day; they just see it at the right time.

Relation to the rest of the framework

  • Upstream: ACT produces the outputs. ACT-API adapts them for consumption.
  • Downstream: care-record systems — e.g. Person Centred Software, the NHS record, a provider dashboard — consume those outputs and surface them in the contexts practitioners already work in.
  • In parallel: the anticipatory.care standalone observational app is a separate route to market for the ACT tool as a whole. ACT-API is the route for MCDC-integrated deployments that need to live inside existing care systems.

References

  • Whalley, J. H., James, L., Brill, M., & Cunningham, S. (2026). Music-Centred Dementia Care: A Dual-Purpose Framework for Therapeutic Intervention and Anticipatory Observation [Manuscript under review]. University for the Creative Arts. See Section 6.
  • ACT observational tool, anticipatory.care — standalone product site with roadmap notes on API integration.