Short answer
A sustainable clinical service can meet the needs of its population, deliver acceptable outcomes and access, maintain the workforce and clinical dependencies on which it relies, and do so within an affordable resource model over time.
Sustainability is not the same as present-day safety, financial balance or regulatory compliance. A service may pass today's tests while depending on a fragile rota, an ageing workforce, cross-subsidy, unreliable clinical dependencies or demand it can no longer absorb. It becomes fragile before it becomes visibly unsafe.
Boards should judge the service as a system. The question is not simply whether it performs adequately now, but whether the relationship between population need, service model, workforce, capacity, dependencies, outcomes and cost can hold over time, and whether a better alternative exists.
Why it matters
Clinical services rarely fail at a single moment. Pressure accumulates. Vacancies become normal, waiting lists lengthen, temporary staffing rises, dependent services compensate and clinical teams spend more time holding the model together. The service may remain compliant, but its margin for absorbing change narrows.
That lag matters. Regulatory indicators often identify visible failure after the operating model has already weakened. Financial reporting can show the cost of pressure without explaining its cause. Workforce reports can show vacancies without testing whether the underlying service model is viable. Each function sees part of the problem.
For a board, sustainability is a forward-looking judgement. It asks whether the service can continue to meet need under plausible changes in demand, workforce, technology, clinical standards and funding. It also asks whether strengthening the current model is the right response, or whether networking, redesign, consolidation or a different pathway would create more value.
What drives sustainability
A useful causal chain is:
The sustainability chain
Weakness at any point can destabilise the rest. Resilience is the result, not an input.
Population need establishes the purpose and likely scale of the service. Historic activity alone is not enough: it reflects what the system supplied, constrained and recorded, not necessarily the need it should meet.
The service model determines how care is organised: what is delivered, where, at what scale, through which pathway and with which relationships to other services. A model designed for yesterday's demand or workforce may remain busy while becoming progressively less viable.
Workforce and capacity determine whether the model can operate safely and reliably. Headcount is only part of the test. Skill mix, senior cover, rota depth, recruitment, retention, succession and dependence on a few individuals all matter. A rota that works only through goodwill or repeated temporary cover is evidence of fragility, even before a formal standard is breached.
Clinical dependencies are equally important. Many services rely on diagnostics, critical care, theatres, pharmacy, pathology, rehabilitation, transport or specialist advice elsewhere. Assessing a service in isolation can produce a false conclusion. Its strength may depend on another service whose own position is weak.
Outcomes, quality, safety and access show what the model produces for patients. These measures need to be read together. A service may report acceptable outcomes while access deteriorates, or maintain throughput by transferring pressure elsewhere in the pathway.
Cost and economic viability show whether the model can be sustained within available resources. The cheapest configuration is not automatically the most sustainable, and a clinically attractive change does not automatically release cash. Boards need to distinguish accounting effects, avoidable cost, investment requirements and value across the whole pathway.
Resilience is the result of these relationships. It is the service's capacity to absorb disruption and adapt without unacceptable deterioration in access, quality, workforce or cost.
Fragility before failure
The most useful assessment identifies the binding constraint before it becomes a crisis.
Consider a small acute service that is currently meeting its quality standards. Demand is stable, but the rota depends on a handful of senior clinicians, recruitment has repeatedly failed and several neighbouring services depend on its availability. The immediate data may not justify calling the service unsafe. Yet one resignation, new professional standard or loss of a dependent service could make the model unworkable.
Recruiting another clinician may relieve the immediate pressure. It does not answer whether the service has sufficient scale, whether the post is recruitable, whether a networked model would be stronger, or how change would affect access and dependent services. Those are questions about the whole system, not one vacancy.
A service becomes fragile before it becomes visibly unsafe. The margin goes first, and the indicators go last.
This is why sustainability cannot be reduced to a red, amber, green dashboard or a single score. The evidence must support a clear conclusion about where the service is strong, where it is fragile, what is driving that fragility and what alternatives are credible.
Better than the alternative
A sustainable service is not simply one that can be kept running. It should also be preferable to the realistic alternatives available to the population and system.
That comparison matters when leaders consider centralisation, networks or service redesign. Centralisation may improve rota depth, concentration of expertise or access to dependencies. It may also increase travel, create pressure at the receiving site, weaken linked services and require investment that does not produce a net saving. The answer depends on the service, geography, population and wider system.
Boards should compare credible options against the same dimensions. These may include retaining and strengthening the existing service, changing its scope, creating a managed network, consolidating selected activity or redesigning the pathway. Doing nothing should be modelled as an active option with consequences, not treated as a neutral baseline.
What leaders should test
A board-level assessment should answer seven practical questions:
- Need: What need should the service meet now and in the future, and for whom?
- Model: Is the current pathway, scale and configuration suited to that need?
- Workforce: Can the required workforce be recruited, retained, deployed and renewed?
- Dependencies: Which other services make this model possible, and how resilient are they?
- Performance: Are outcomes, quality, safety and access acceptable, and in which direction are they moving?
- Economics: Is the model affordable, and what value does it create across the pathway rather than within one budget?
- Alternatives: Is this configuration better than realistic alternatives, including the consequences of delay?
The test should expose uncertainty rather than hide it. Leaders need to know which conclusion is supported directly by evidence, which depends on assumptions and which would change if a key constraint moved. That is a Decision Intelligence problem: the whole decision has to be reconstructed from evidence that organisations usually hold in separate functions.
The Strasys approach
We assess clinical services through the Clinical Service Evaluation System, using ten dimensions organised across four stages: Demand, Inputs, Outputs and Context.
Four stages, ten dimensions
Two cross-cutting overlays, time and change, and equity, apply across every stage.
Demand
Why and how much work?
- Population need
- Activity volume
Inputs
Can it be delivered?
- Workforce
- Clinical dependencies
- Estate, equipment and digital
Outputs
Is what it produces acceptable?
- Outcomes, quality and safety
- Cost and economic viability
Context
Does it sit in a viable, resilient system?
- Geography and alternative locations
- Resilience and concentration
- Governance, partners and enablers
Six of the ten dimensions have quantitative models. The remaining four combine quantitative inputs with structured qualitative assessment. See the interactive Diagnostic Wheel for the full method.
The evidence is brought together on a one-page Diagnostic Wheel. The wheel is not intended to disguise complexity in a composite score. It shows the pattern of strength and fragility across the service so leaders can see the whole decision and identify the binding constraints.
The system supports repeatable portfolio assessment before difficulty sets in. A Clinical Service Review serves a different purpose: it brings structured, objective analysis to a specific service or decision already requiring attention. The approaches should link, but they should not be presented as interchangeable.
In practice, the assessment can inform reconfiguration, network design, succession planning, capital prioritisation, response to regulatory concern and wider clinical strategy. The conclusion should connect evidence to a decision: what should be protected, what should change, what the alternatives would produce and what leaders need to monitor.
Evidence and limitations
External guidance supports examining clinical evidence, workforce, interdependencies, whole-system consequences, deliverability and affordability when significant service change is considered.1 It does not provide a universal formula for sustainability, nor does it justify a default assumption that larger or more centralised services always produce better outcomes or lower costs.
The definition and causal chain in this article are our interpretations, developed for healthcare-management decisions. The Clinical Service Evaluation System is the formal method we use to structure the assessment.2 Our public Blackpool cardiology and cardiothoracic case supports application of the review method across services serving a reported population of 1.7 million.3 Realised outcomes from that engagement are confidential and are not claimed here.
A sustainability assessment is constrained by the quality and comparability of available evidence. Some services have strong national standards and outcome data; others require calibrated proxies and professional judgement. Where data is incomplete, the uncertainty and its importance to the decision should be explicit.
A service can also be sustainable for an organisation while producing poor value for the wider system or population. The unit of analysis must therefore be clear. Provider viability, pathway value and population benefit are related, but they are not the same conclusion.