Why some older mental health wards look expensive compared to benchmark

A question comes up regularly in mental health inpatient services:

“Why is this ward so expensive compared to benchmark?”

On paper, it often looks straightforward. Bed-day costs are high. Staffing costs look heavy. Bank usage may be elevated.

And to be clear, the costs are high.

But in older mental health inpatient settings, high cost does not automatically mean poor operational management or inefficiency. Often, it reflects something more structural: the ward is being run safely within the constraints of the physical environment.

That distinction matters, because it changes the question leaders need to ask.

This issue tends to show up in older estate

This is not a universal point about all inpatient services.

It comes up most often in:

• mental health inpatient services
• particularly learning disability and autism settings
• wards operating from older, constrained, or poorly configured estate

If a service is running from a modern, purpose-built environment, the issue may be less pronounced.

But where wards are operating from older buildings, split layouts, or environments not designed for current acuity, the relationship between cost, staffing and occupancy often looks very different.

What benchmarks miss

Benchmarks do not usually state an explicit occupancy assumption.

But in practice, they are based on services that are typically able to:

• run at relatively high and stable occupancy
• operate in layouts that support safer observation
• use staffing models that flex more easily with demand

In other words, benchmarks tend to reflect how services perform when they can run closer to full and do so safely.

That is important, because some older wards simply cannot operate at that level.

The more useful question

So instead of starting with:

“Why is this ward so expensive?”

a better starting point is:

“What level of occupancy can this ward actually accommodate safely?”

Because if a ward cannot safely achieve the sort of utilisation seen in benchmark services, cost per bed-day will look high almost regardless of how well the team is managing the ward.

What we often see in practice

In one recent piece of work, we analysed a group of mental health inpatient wards where headline benchmark comparisons suggested costs should have been closer to peer levels.

But the operational reality looked different.

At ward level:

• typical occupancy was around 8 beds
• designed capacity was 10 to 12 beds

That gap mattered.

It meant cost per patient looked high, but not simply because the service was being run inefficiently. The bigger issue was that safe operating capacity was materially lower than theoretical capacity.

Why safe capacity was lower than designed capacity?

Three issues kept surfacing.

1. Layout

Older wards often had poor sightlines, segmented spaces and layouts that made observation harder.

That meant adding patients did not automatically improve efficiency. In some cases, it simply increased operational risk.

2. Cohorting and escalation (particularly in LD&A)

In learning disability and autism settings, the environment can directly influence behaviour.

• Patients may be less able to tolerate busy or poorly structured environments
• Escalation in one individual can trigger escalation in others
• Layout can make de-escalation harder to manage safely

👉 In practice, this often reduces the number of patients that can be safely accommodated at any one time

3. Staffing dynamics

Staffing does not behave as either fully fixed or fully variable.

• There is a baseline level of staffing required to operate safely
• Above that, staffing can flex with acuity, observations and risk
• In more constrained environments, higher staffing levels are often needed to manage risk

👉 So even where occupancy is lower, costs do not reduce proportionately

Why the numbers then look so poor?

Once that happens, the maths becomes challenging.

If:

• a significant proportion of cost is relatively fixed
• staffing flex is limited by risk and environment
• occupancy is constrained for safety reasons

then:

cost per bed-day rises

That does not prove inefficiency. It may reflect the fact that the ward cannot operate at the level that benchmark services are able to achieve.

Why this matters for decision-making?

If leaders treat this purely as a cost problem, the response is usually predictable:

• reduce staffing
• challenge bank usage
• push for better productivity

There may well be opportunities in those areas.

But if the core issue is constrained safe capacity driven by the estate and operating model, those actions will only go so far.

They do not change the underlying operating reality.

What can be done?

This is not an argument for inaction.

But the most effective responses tend to focus on the underlying drivers, not just the symptoms.

That might include:

• improving visibility and observation within existing layouts
• rethinking how patients are cohorted and managed across spaces
• making targeted changes to how staffing is deployed across the ward
• and, in some cases, considering more fundamental changes to the physical environment or configuration of services

The key is to align the operating model with what the environment can safely support — or change the environment.

What leaders should test first?

Before concluding that a ward is inefficient, it is worth testing three things:

1. What is the ward’s true safe operating capacity, not just its designed capacity?
2. How far is the physical environment constraining observation, cohorting and flexibility?
3. Are comparisons being made against a realistic peer group, or against services operating in materially different conditions?

Those questions usually lead to a more useful conversation than cost alone.

Final thought

High-cost inpatient wards should absolutely be scrutinised.

But in older mental health estate, cost, staffing and occupancy are often tightly linked to the physical environment.

If those constraints are not recognised, there is a real risk of diagnosing the wrong problem.

And once that happens, improvement efforts can become focused on the symptoms rather than the cause.