Skip to content

Edit this page

FHIR research notes for digital-pews

Date: 2026-04-21 Status: working notes for local design and adapter planning

Research question

How confident are we that the current FHIR chart adapter spec is correct, and has PEWS already been modelled in FHIR?

Sources reviewed

  1. NHS England / NHS Digital API catalogue

  2. https://digital.nhs.uk/developer/api-catalogue

  3. Outcome: no clearly discoverable PEWS-specific production API entry from this route.

  4. Simplifier project search

  5. https://simplifier.net/search?term=PEWS&scope=UK

  6. Outcome: surfaced relevant UK discovery/draft artifacts including Clinical Observations and a separate PEWS logical model page.

  7. NHS Digital Clinical Observations IG (R4)

  8. https://simplifier.net/clinicalobservations

  9. https://simplifier.net/guide/clinicalobservations

  10. Key status text: draft/discovery guidance under active development.

  11. Clinical Observations examples page/toc

  12. https://simplifier.net/guide/clinicalobservations/home-examples-toc

  13. Outcome: includes PEWS and NPEWS1 examples, including bundle examples and score examples.

  14. Example pages found in the Clinical Observations guide

  15. https://simplifier.net/guide/clinicalobservations/PEWSTotalScoreWithObservationsBundleExample

  16. https://simplifier.net/guide/clinicalobservations/BaselineNPEWS1ObservationExample
  17. https://simplifier.net/guide/clinicalobservations/NPEWS1WithSubScoresObservationExample
  18. https://simplifier.net/guide/clinicalobservations/ScoreChartUsedExample

  19. https://simplifier.net/guide/clinicalobservations/RespiratoryDistressObservationExample

  20. Separate PEWS logical model page (non-NHS formal profile set)

  21. https://simplifier.net/ckmfhirlogicalmodel/pews-originalvariables

  22. Outcome: appears to be a draft logical model artifact, useful for ideas but not authoritative for implementation.

Findings

  1. PEWS has been modelled in FHIR contexts already.

  2. There is NHS Digital/NHS England Clinical Observations work in FHIR R4 that explicitly includes early warning score patterns and PEWS examples.

  3. There are example artifacts for PEWS total score bundles and NPEWS1 observations with sub-scores.

  4. The national-state signal is still draft/discovery.

  5. The Clinical Observations guide indicates active development and draft/discovery state.

  6. It should be treated as a strong directional reference, not a final stable national conformance target.

  7. The existing local adapter spec direction is broadly sound.

  8. Bundle-oriented event exchange, Observation-centric vitals, and separate score semantics are aligned with the direction of the Clinical Observations content.

  9. Existing local assumptions about using standard Observation patterns and absent-data handling are compatible with mainstream FHIR design.

  10. Important nuance for scoring representation.

  11. The Clinical Observations examples indicate sub-score representation patterns that should be reviewed against local triggerCodes and score decomposition assumptions.

  12. Local model fields should remain flexible enough to map to component-based score details.

Confidence rating

  • Confidence that PEWS has prior FHIR modelling work: high.
  • Confidence that there is a fully stable, nationally fixed PEWS implementation profile set ready for strict conformance: medium-low.
  • Confidence that our current adapter design approach is directionally correct: medium-high.

Implications for this repository

  1. Keep the local canonical model and adapter boundary.

  2. The local chart model remains useful and should not be forced to mirror raw FHIR resources.

  3. Treat Clinical Observations as the primary external reference.

  4. Use it to guide coding systems, score decomposition shape, and example payload design.

  5. Preserve local flexibility in unresolved areas.

  6. Keep extension-friendly handling for PEWS-specific semantics where national guidance remains draft.

  7. Add explicit test strategy for score consistency.

  8. Continue carrying expected scores in test fixtures as assertions.

  9. Validate that recomputed score totals and sub-scores match expected values for each age band.

Suggested next steps

  1. In spec/fhir.md (future-work appendix), keep explicit references to the Clinical Observations IG and caveat text about draft/discovery status.
  2. In spec/fhir.md, add a short compatibility note on score component mapping from NPEWS1 examples.
  3. Add adapter conformance tests that compare chart-model score fields against known PEWS/NPEWS fixture expectations.
  4. Keep this file as the running research log until a consolidated docs/ structure is introduced.

Future documentation structure (proposed)

When moving to a docs site, consider this information architecture:

  • docs/spec/
  • local architecture and adapter contracts
  • docs/reference/
  • external standards summaries and links (including FHIR research)
  • docs/research/
  • exploratory notes, option analysis, unresolved decisions
  • docs/how-to/
  • practical guides for developers and clinical reviewers

This note can migrate to docs/reference/fhir-research.md when docs are consolidated.