ISLANDS OBSERVATORYhealthcare
03 · Governance — reviews, inspections and audit: did anyone do what they said?
Public lensEvery few years your health system gets reviewed or inspected. The findings make headlines; the follow-through rarely does. This tracker holds every recommendation we could extract — with its status where one has ever been published, and an honest Unknown where it hasn't.
Politician lensThe scoreboard shows who tracks implementation and who doesn't. Below it, filter your jurisdiction's stack to Identified / Accepted / Unknown — each is a question waiting to be asked. IoM items carry ⚑ objection flags with Hansard links; the question-generator lives on each card via the Hub's Focus panel.
Management lensPractice worth borrowing: Jersey implemented a structured recommendations-tracking system (C&AG R21, verified 2022) and publishes C&AG follow-ups that re-grade every rec. Scotland's HIS inspects on a rolling cycle with published improvement plans. IoM statuses are reconstructable from Hansard. Guernsey has none of these — the cheapest upgrade on this page.
Direct numerical comparison without reading caveats is misleading. Each jurisdiction defines, measures and reports on its own basis — hover or tap any figure for its basis and source before drawing conclusions.
Cross-island scoreboard
Review timeline — click a dot to jump · dim dots = stubs / FOI pending
Status
Impact
Methodology & data governance notes
Sources. Every displayed figure carries a source string (document + date) reachable by hover, tap or inline chip. The consolidated data layer (observatory_data.js) is generated by script from extraction CSVs in the observatory library; nothing is hand-typed into pages. Primary sources: government/board annual reports and accounts, Audit Scotland/KPMG annual audits, HIS inspection reports, C&AG/JAO reports, DHSC Assessments of Manx Care, Tynwald Hansard and written answers, ONS/NRS/PHS statistics, CAA punctuality statistics, and FOI responses held in the library.

Comparability. Jurisdictions differ in definitions, targets, reporting windows and population bases. Basis lines are mandatory on comparison rows; 3-year rolling averages (small populations) are marked and never silently mixed with single-year rates; derived figures (per-capita, per-1,000, inversions) are labelled derived; DERIVED Sankey nodes state their derivation.

Gaps. Missing data is displayed, not hidden: red chips = no comparable public figure exists; blue chips = published somewhere we have not yet extracted (each maps to a roadmap item). Gaps are findings.

Views. The Public view uses plain language and does not draft formal requests; Politician and Management views expose question/FOI wording and benchmark framing. The view switch changes emphasis, never the underlying numbers.

Corrections. Errors: observatory@coalfinch.com. Build log and change history: see the Roadmap page.