Healthcare doesn't move in straight lines. It pulses through wards, across organisations, into homes, and back again. It involves split-second decisions and decade-long continuity. It depends on people who may never meet but must still trust each other's judgement.
Behind all of this movement is information — not as a passive record, but as a living current. And like any complex system, healthcare depends on certain flows behaving predictably. When they do, care feels seamless. When they don't, the consequences can be profound.
This essay describes seven foundational flows of care: Identity, Consent, Provenance, Clinical Intent, Alert and Responsibility, Service Routing, and Outcome. Together, they form the connective tissue of modern health systems — enabling trust, continuity, safety, and learning across time and space.
Each flow is examined through the lens of 4W1H: Why it matters, What it is, Who it involves, When it occurs, and How it operates. This structure helps us see each flow not as an abstraction but as something enacted — by people, in moments, under pressure.
1. Identity Flow
"Before anything else, we must know who we're caring for."
Why: Safe care depends on matching the right information to the right person. Misidentification can lead to wrong treatments, missed allergies, or fragmented records. Every downstream decision — from drug dosing to discharge planning — rests on knowing who the patient is.
What: Identity Flow is the continuous process of establishing, verifying, and linking a patient's identity across encounters, systems, and time. It includes demographic matching, NHS number validation, and cross-referencing with the Personal Demographics Service (PDS).
Who: Front-line staff — receptionists, nurses, paramedics — often initiate identity verification. Behind them, system integrators, registration authorities, and data stewards maintain the infrastructure. Patients themselves participate through ID checks, app logins, and confirming their details.
When: Identity is verified at first contact and re-confirmed at key moments: admission, handover, prescribing, results review. In digital systems, identity tokens persist across sessions, but human confirmation remains essential at safety-critical points.
How: The NHS Number serves as the primary identifier, supported by PDS lookups and local Master Patient Indexes (MPIs). FHIR Patient resources standardise how identity is represented. In practice, probabilistic matching algorithms handle variations in spelling, date formats, and aliases. Two-factor verification and biometric options are increasingly used in patient-facing apps.
2. Consent Flow
"Trust is given, not assumed."
Why: Consent is the ethical and legal foundation of data use in healthcare. It reflects the patient's right to control how their information is shared. Without clear, dynamic consent, organisations risk both legal breach and erosion of public trust.
What: Consent Flow captures the patient's preferences about data sharing — who can see what, under which circumstances. It includes explicit opt-ins, implied consent for direct care, and the ability to withdraw or modify preferences over time.
Who: Patients are the primary actors. Clinicians and care coordinators must respect consent boundaries. Data controllers and processors operationalise consent logic, while governance teams audit compliance.
When: Consent is captured at registration, updated during care episodes, and revisited when data use changes — for example, when information is shared for research or secondary purposes. It must be queryable in real time.
How: Consent is managed through systems like the National Data Opt-Out, local consent registries, and FHIR Consent resources. Policies are encoded as machine-readable rules that gate access at the point of query. Audit trails ensure that consent decisions are traceable and defensible.
3. Provenance Flow
"Every piece of information has a story."
Why: In healthcare, context is everything. A blood pressure reading means something different if taken at rest versus mid-exertion, by a clinician versus a home device. Provenance tells us where data came from, how it was captured, and whether it can be trusted.
What: Provenance Flow tracks the origin, transformation, and movement of clinical data. It includes who created a record, when, using what device or method, and any subsequent changes or derivations.
Who: Clinicians, devices, algorithms, and patients all generate data. Integration engines and middleware transform it. Data stewards and archivists preserve provenance metadata. Auditors and regulators may later interrogate it.
When: Provenance is recorded at the point of data creation and updated with each transformation or transfer. It should persist for the life of the record — and often beyond, for medico-legal and research purposes.
How: FHIR Provenance resources capture authorship, timestamps, and source systems. W3C PROV standards offer a richer model for complex data lineage. Integration platforms like Apache NiFi or Microsoft Azure Data Factory can embed provenance into data pipelines. Immutable logs and blockchain-inspired approaches are emerging for high-assurance contexts.
4. Clinical Intent Flow
"It's not just what was done, but what was meant."
Why: Healthcare involves decisions under uncertainty. Understanding the reasoning behind an action — not just the action itself — is essential for continuity, safety, and learning. Clinical intent bridges the gap between raw data and meaningful care.
What: Clinical Intent Flow captures the purpose behind clinical actions: why a medication was chosen, what a referral was meant to achieve, what outcome was anticipated. It includes care plans, problem lists, goals, and the narrative reasoning documented by clinicians.
Who: Clinicians are the primary authors of intent. Patients contribute through shared decision-making. Care coordinators synthesise intent across providers. Decision support systems may infer or suggest intent based on patterns.
When: Intent is documented at the point of decision — during consultations, multidisciplinary meetings, or care planning sessions. It may evolve as conditions change and should be reviewed at transitions of care.
How: Structured fields like FHIR CarePlan, Goal, and Condition resources encode intent. Free-text notes remain important for nuance. Natural language processing (NLP) is increasingly used to extract and codify intent from narrative. Clinical decision support tools can surface relevant intent at the point of care.
5. Alert and Responsibility Flow
"When something matters, the right person must know."
Why: Healthcare generates a constant stream of signals — vital signs, lab results, risk scores. Not all require immediate action, but some do. The ability to surface the right alert to the right person at the right time is a matter of safety.
What: Alert and Responsibility Flow manages the generation, routing, and acknowledgement of clinical alerts. It includes threshold-based triggers, escalation rules, and mechanisms for confirming that a responsible party has received and acted on an alert.
Who: Alerts may originate from devices, algorithms, or human observers. They are routed to clinicians, nurses, or on-call teams based on responsibility models. Patients and carers may also receive alerts in remote monitoring scenarios.
When: Alerts are generated in real time or near-real time. They must be delivered promptly, with escalation if unacknowledged. Some alerts are time-bounded; others persist until resolved.
How: Clinical alerting systems integrate with EPRs, device gateways, and communication platforms. FHIR Flag and CommunicationRequest resources can represent alerts. Responsibility is often managed through rota systems, role-based routing, and escalation trees. Closed-loop acknowledgement ensures accountability.
6. Service Routing Flow
"The right care, in the right place, at the right time."
Why: Modern healthcare is distributed across providers, settings, and geographies. Getting a patient to the right service — whether a specialist clinic, a diagnostic facility, or a community team — requires coordination and visibility.
What: Service Routing Flow governs how patients are directed through the system. It includes referrals, bookings, triage decisions, and capacity management. It ensures that demand meets supply efficiently and equitably.
Who: GPs, clinicians, and care coordinators initiate routing decisions. Booking and scheduling teams operationalise them. System managers monitor flow and capacity. Patients navigate the system with varying degrees of agency.
When: Routing decisions occur at transitions: referral from primary to secondary care, discharge planning, follow-up scheduling. Real-time routing is increasingly important in urgent and emergency care.
How: E-Referral systems, directory of services (DoS), and appointment booking APIs enable routing. FHIR ServiceRequest, Appointment, and Slot resources standardise representation. Intelligent routing algorithms can match patient needs to service capabilities, considering geography, wait times, and clinical criteria.
7. Outcome Flow
"What happened — and what did it mean?"
Why: Healthcare exists to improve health. Without capturing outcomes, we cannot know whether we succeeded. Outcome data closes the loop — informing individual care, population health, research, and system improvement.
What: Outcome Flow captures the results of care: clinical outcomes (e.g., recovery, complication), patient-reported outcomes (e.g., quality of life, symptom burden), and process outcomes (e.g., length of stay, readmission). It links back to the intent and actions that preceded it.
Who: Clinicians record clinical outcomes. Patients report their experience and functional status. Analysts aggregate outcomes for benchmarking and research. Quality teams use outcomes to drive improvement.
When: Outcomes are measured at defined endpoints: post-procedure, at discharge, at follow-up intervals. Some outcomes emerge over months or years and require longitudinal tracking.
How: Structured outcome measures — PROMs, PREMs, clinical endpoints — are captured through EPRs, patient apps, and registries. FHIR Observation and QuestionnaireResponse resources can encode outcomes. Linkage to intent (via CarePlan or Goal) enables outcome-based evaluation of care effectiveness.
Seeing the System Whole
These seven flows are not separate systems — they are interdependent. Identity enables Consent to be applied. Provenance gives meaning to Clinical Intent. Alerts depend on Service Routing to reach the right responder. Outcomes only make sense when linked back through the chain.
When these flows are well-designed and well-governed, healthcare becomes more than a collection of transactions. It becomes a coherent system — one that learns, adapts, and earns trust.
At Inference Clinical, we work with NHS organisations and innovators to strengthen these flows. Whether through clinical safety assurance, interoperability design, or outcome-focused strategy, our goal is the same: to help care flow as it should.
