Skip to content
Healthcare operations

Operational quality, investigated with sources. Measures and policy, together.

We keep the definition in force attached to every operational number. Our work here is operational and administrative only.

Knowledge factory. A knowledge factory line: Documents, Tables, Conversations, Systems → Ingest → Extract → Resolve → Map → Validate → Publish → Serve. Items that fail a constraint go to a review queue, where a person approves or rejects them. So far: 11 ingested, 22 extracted, 7 merged as duplicates, 12 passed validate, 1 rejected, 13 published, 11 served as context. The followed item (term “late delivery”) is at: Map.

Documents and tables become reviewed knowledge before anything uses them. Illustrative data.
What is hard here

Moving definitions.

Operational measures move when their definitions change. Reports rarely say which happened.

What we have learned

Half of a measure’s movement can come from a new specification. Pinning the version in force for each period separates real change from a change in the rule.

Agent Hive · Run 5120
Why did our discharge-summary turnaround measure slip this quarter?
  1. Checkedrequester role · quality-analyst0.2s
  2. Readmeasure spec v7 and v8 · policy QM-141.0s
  3. Queried9 sites · 41,200 encounters2.2s
  4. RecomputedQ3 under v7 for comparison1.1s
  5. Wrotefindings with 4 sources0.3s
Answer4 sources
−6.1 pts78.4%
Summaries within 48 h, Q3

About half the drop is the v8 definition change to the 48-hour window. The rest comes from 2 of 9 sites after a scheduling change.

measure spec v8 · p.3policy QM-14encounters_opssite_rosters
Ask a follow-up…
Illustrative data
Figures

Measures, with their versions.

Fig. 1 · Definitions change through approvalIllustrative data

Ontology evolution. How the structure changed over 4 proposals: v1: baseline with Thing, Document, Report, Metric, Party, Vendor, Supplier. P-1 (add, from steward): Add class Definition ⊑ Metric, applied, giving v2. Diff: + Definition subClassOf Metric. P-2 (merge, from resolver): Merge Vendor into Supplier, applied, giving v3. Diff: − Vendor; + Supplier altLabel “Vendor”. P-3 (add, from extractor): Add class Misc ⊑ Document, rejected; v3 unchanged. Diff: + Misc subClassOf Document; ✗ rejected: no defining property. P-4 (deprecate, from steward): Deprecate Report, applied, giving v4. Diff: ~ Report owl:deprecated true; + replaced by Document. Selected: P-4 (deprecate, from steward): Deprecate Report, applied, giving v4. Diff: ~ Report owl:deprecated true; + replaced by Document.

Proposal
Each measure definition is a version. The analysis uses the one in force for the period.
Fig. 2 · One model of the organisationIllustrative data

Ontology graph. An organisation ontology: 13 classes in an is-a tree and 11 instances. In focus: Metric is-a Measure. A measure computed from data, defined by a term. Instances: cycle time, on-time rate. defined-by Term; relates-to (computed from) Dataset.

Focus
Processes, measures, policies and roles in one ontology.
Fig. 3 · Inside your boundaryIllustrative data
Air-gapped deploymentEverything runs inside the customer’s perimeter with no outbound internet: users, the agent runtime, local models, the knowledge graph, the audit log and read-only connectors to on-prem systems. Updates come in as signed offline bundles. Point at a part to see what it holds and what is blocked.OUTSIDEReleasesigned bundlevX.Y · checksumcarried on offline mediano network linkINTERNEThosted model APIsCUSTOMER PERIMETERNO EGRESSUsersSSO · RBACcustomer identityAGENT HIVE RUNTIMEAgents · workflowsApproval gatesPoliciesEvals · tracesLocal modelsGPU servinginside the siteKnowledge graphdocuments + systemsstored locallyAudit logwho · what · whenevery approvalON-PREM SYSTEMSREAD-ONLY CONNECTORSERPData lakeDocumentsagents read through approved, read-only accessREADUpdate importverify signatureadmin approvesAPPROVED UPGRADEno outbounddata, models, graph and audit never leave the line

Point at a part no egress · signed bundles in · data stays inside

Everything runs on site, connectors are read-only and records never leave.

Bring us one hard problem. We will say whether it fits our research.