A digital health tool can save time without saving the NHS money

A digital tool saves two minutes during an appointment. Multiply two minutes by thousands of appointments, attach a salary cost, and the business case appears to have found a large saving.

The arithmetic may be flawless. The saving may still not exist in a budget.

Staff remain employed. The clinic still opens for the same hours. The software has a licence fee, requires support, and may move work to another team. The two minutes could be valuable, but “valuable” and “cash-releasing” are different claims.

Time saved is capacity released. What that capacity becomes is a service design and economic question.

This distinction is useful when reading almost any health-technology proposal. It does not argue against investment. It makes the case describe the benefit that can actually be delivered.

Start with the counterfactual

Economic value is comparative. The relevant question is not “does the technology have benefits?” but “what changes compared with what would otherwise happen?”

The counterfactual might be:

the existing manual process;
a smaller change to the current system;
a competing product;
additional staff or equipment;
no intervention while demand continues to change.

If the baseline quietly assumes that every current problem persists while the new technology works perfectly, the comparison has already chosen a winner.

Document the current pathway with the same care as the proposed one. Include waiting, duplication, workarounds, error correction, escalation, and the people whose effort is easy to overlook.

Separate four kinds of value

“Saving” often bundles several outcomes that need different evidence.

Cash-releasing savings

These reduce actual expenditure: a contract ends, paid overtime falls, agency usage drops, or a budgeted post is no longer required.

Cash release usually needs an operational action beyond installing software. A few minutes distributed across many staff do not become cash unless work and resources can be reorganised.

Capacity

Released time can allow more appointments, faster processing, shorter queues, or more attention for complex work. This may be the most important benefit in a constrained service even when expenditure does not fall.

Capacity should be described in service units where possible: additional appointments, reports, referrals, calls, or cases at the relevant quality.

Quality and outcomes

A technology may improve consistency, access, safety, patient experience, staff experience, or clinical outcomes. These benefits do not need to masquerade as cash savings to matter.

They do need appropriate evidence. A better interface is not automatically a better outcome, and higher usage is not automatically improved health.

Cost displacement

Work or cost may move rather than disappear. Patient self-entry can reduce administration while increasing the burden on patients. Automated triage can reduce one queue while creating additional review or escalation elsewhere.

Always ask who receives the benefit and who takes on the new work.

The two-minute example

Consider a deliberately simple illustration:

20,000 appointments × 2 minutes
= 40,000 minutes
= approximately 667 hours

The calculation estimates gross time released during the measured period. It does not yet estimate a financial saving.

Those 667 hours might become:

additional patient-facing capacity;
less overtime;
reduced backlog;
more complete documentation;
shorter appointments;
unused fragments of time that cannot be scheduled;
extra checking required because the new process creates different risks.

To claim a cash saving, the model needs a credible route from small increments of released time to a change in expenditure. To claim capacity, it needs to show that the time can be assembled, scheduled, and used. To claim quality, it needs an outcome measure.

The same time figure can support several stories. The operating model decides which story is real.

Count implementation as part of the technology

The purchase price is rarely the full cost of adoption.

A digital health implementation may require:

workflow design and clinical safety work;
integration, data migration, and identity configuration;
information governance, data protection, and cyber-security assurance;
accessibility and usability assessment;
staff training and backfill;
patient support and assisted-digital routes;
hardware, connectivity, and local infrastructure;
supplier management, monitoring, incident response, and upgrades;
parallel running, evaluation, and eventual decommissioning.

NHS England’s Digital Technology Assessment Criteria cover clinical safety, data protection, technical security, interoperability, usability, and accessibility. Those are not decorative procurement headings. They point to real work required to deploy and operate technology safely.

The cost profile also changes over time. Integration is not a one-off if interfaces, suppliers, regulations, or surrounding services change.

Measure adoption, not licences

A licence can be deployed without the intended behaviour becoming routine.

Economic benefit depends on a chain:

The technology is available.
Staff and patients can use it.
The relevant workflow actually changes.
The change affects time, quality, access, or outcomes.
The service converts that effect into a valuable result.

Failure at any link weakens the downstream claim.

Measure usage by the people and situations for which the benefit was predicted. An overall adoption percentage can hide that the tool works well in simple cases but is bypassed when the workload is complex—or that it excludes people who need a different route.

Choose the perspective before counting

A change can save money for one organisation and increase cost elsewhere.

A hospital may discharge earlier while community services or families take on more work. A digital front door may reduce telephone demand while creating a new review queue. Remote monitoring may avoid travel for some patients while requiring devices, connectivity, and support.

HM Treasury’s Green Book asks appraisals to look beyond the originating organisation and consider significant costs, benefits, and risks for other public bodies, businesses, households, and charities. It also distinguishes public-sector financial effects from wider social value.

State the perspective:

a team or clinic;
an NHS provider;
an integrated care system;
the public sector;
patients, carers, and society.

There may be good reasons to present several. Do not add them together without checking whether the same benefit has been counted twice.

Cost-effective does not mean cheaper

An intervention can represent good value while increasing total spending.

If a technology improves outcomes enough to justify its additional cost, it may be cost-effective without being cost-saving. Conversely, a tool can reduce one budget while worsening outcomes or transferring cost, making it a poor choice from a wider perspective.

NICE’s Evidence Standards Framework for digital health technologies explicitly separates evidence of performance from evidence of economic impact. That is a helpful discipline: first establish what the technology does, then model what that change is worth under realistic adoption and cost assumptions.

Make uncertainty do some work

A business case contains assumptions about adoption, time saved, staff mix, implementation duration, recurring cost, demand, and outcomes. A single central estimate can make those assumptions disappear.

Use ranges and sensitivity analysis:

What if adoption reaches 40% rather than 80%?
What if each case saves 30 seconds rather than two minutes?
Which implementation costs recur?
At what workload does the investment stop representing value?
Which assumption changes the conclusion first?

The Green Book recommends sensitivity analysis and explicit adjustment for optimism bias. The point is not to make innovation look unattractive. It is to identify the conditions required for the proposed value to exist.

A better claim

Replace:

This tool saves 667 staff hours and therefore £X.

With something closer to:

At 20,000 eligible appointments, the observed two-minute workflow reduction would release approximately 667 staff hours. The implementation plan will consolidate that time into additional appointment capacity. The economic model includes training, integration, support, and an adoption range of 40–80%; no cash-releasing saving is assumed.

That claim is less dramatic. It is also easier to test after deployment.

The useful insight

Digital health economics improves when “saving time” is treated as the beginning of the argument rather than the conclusion.

Follow the released minutes. See whether they combine into usable capacity, change expenditure, improve outcomes, or move work to somebody else. The value may be substantial—but it should be named accurately.