NHS · DSPT 2026-27 · Independent audit
DSPT audit 2026-27: what the independent assessor checks, and how to pass
Every CAF-aligned organisation needs its Data Security and Protection Toolkit independently audited before it publishes. This page covers the audit itself: the eleven mandatory outcomes plus the one you choose, the twelve areas for IT suppliers, how the assessor tests, how the risk and confidence ratings are decided, and a free check of where you stand.
At a glance
Who needs a DSPT audit
For 2026-27, NHS England has set mandatory audit areas for NHS trusts, integrated care boards, arm's length bodies, commissioning support units, independent operators of essential services and the genomics organisations nominated by DHSC. These organisations complete the CAF-aligned version of the toolkit and have a defined set of outcomes independently assessed before they publish.
NHS England has also published mandatory audit areas for IT suppliers, which use the standards-based toolkit rather than the CAF-aligned one. If you are a GP practice, care provider or other organisation completing the standards-based toolkit, check your own requirements on the toolkit; most do not need an independent audit.
The audit sits alongside your self-assessment. It does not replace it, and its main question is whether your self-assessment can be relied on.
The 2026-27 audit scope: eleven outcomes plus one
NHS England published the 2026-27 mandatory audit areas on 1 October 2026. The same eleven outcomes apply to trusts, ICBs, ALBs, CSUs, OES and genomics organisations, and each organisation adds one outcome of its own choice, agreed with its assessor based on its risks.
| Outcome | What it covers |
|---|---|
| A1.b | Roles and responsibilities |
| A3.a | Asset management |
| B2.c | Privileged user management |
| B2.d | Identity and access management |
| B3.a | Understanding data |
| B4.c | Secure management |
| B6.a | Culture |
| B6.b | Training |
| C1.e | Personnel skills for monitoring and detection |
| D1.c | Testing and exercising |
| E4.a | Managing records |
| +1 | One outcome you choose. NHS England recommends D1.a Response plan for trusts, ICBs, ALBs and CSUs; B5.c Backups for OES and genomics; A2.a Risk management process for CNI operators. |
The scope changes every year. In 2025-26 trusts and ICBs had nine mandatory outcomes plus three of their choice. Check the toolkit for any update before you agree your terms of reference.
IT suppliers: the twelve audit areas
For IT suppliers, NHS England has named twelve areas from the standards-based toolkit for audit in 2026-27. Supplier audit guidance was to be shared separately.
| Area | What it covers |
|---|---|
| 1.1 | Lawfulness, fairness and transparency framework |
| 2.2 | Staff contract responsibilities for data security |
| 3.1 | Staff understanding of information governance and cyber security |
| 3.2 | Proactive engagement and an open and just culture for data security incidents |
| 4.1 | Current record of staff and their roles |
| 6.2 | Anti-virus protection and email spam filtering |
| 7.1 | Defined response to data security incidents |
| 8.1 | Survey of software and hardware support status |
| 8.2 | Categorised and documented unsupported software and hardware |
| 9.2 | Penetration test scoped and undertaken |
| 9.6 | Well-managed firewall |
| 10.2 | Basic due diligence on suppliers handling personal information |
If you sell to the NHS, these are the controls a buyer's due diligence is most likely to probe as well. Our NHS supplier evidence review works through them with you before anyone else does.
How the assessor tests
NHS England's guide for CAF-aligned DSPT independent assessors describes four testing techniques, usually used in combination:
- Inquiry: asking the people responsible how a control works. It gives the least assurance on its own.
- Observation: watching the control operate.
- Inspection: examining documents, records and configurations.
- Automated testing or system query: interrogating the systems directly, which can give the most assurance.
Assessors can test whether a control is well designed, whether it operates in practice, or both. They sample rather than test every instance, with the sample size set by the size of the population and the risk of the control. The guide asks them to judge whether you meet the outcome, not to tick rules, and to use professional judgement about your environment.
In practice, a policy alone rarely satisfies an assessor. Expect to show the control working: the leaver whose account was disabled, the admin account review that removed access, the exercise whose actions were closed.
How you are rated
Each outcome in scope is rated Achieved, Partially achieved or Not achieved against the indicators of good practice, and compared with the minimum level your organisation profile expects. The assessor then gives two overall ratings.
Overall risk rating
| Overall risk rating | When it applies |
|---|---|
| Very low | Every minimum achievement level met, and at least one exceeded |
| Low | Every minimum achievement level met |
| Moderate | No more than one outcome below its minimum level |
| High | Two to four outcomes below their minimum level |
| Very high | More than four outcomes below minimum, and/or mandatory policy requirements cannot be met |
Confidence rating
| Confidence rating | What it means |
|---|---|
| High | Little or no difference between your self-assessment and what the assessor found |
| Medium | A noticeable difference, judged by the assessor |
| Low | A large difference, for example declaring Standards Met when several outcomes fall below minimum |
The confidence rating is the one boards underestimate. It measures how far your own self-assessment can be trusted. Strong scores with low confidence tell NHS England that your published position is not reliable, and that is worse than an honest Partially achieved.
These ratings are from NHS England's guide for CAF-aligned DSPT independent assessors, last updated 9 April 2026 for 2025-26. NHS England said updated guidance for 2026-27 would follow in October 2026; check the toolkit for the current version.
The timetable
Under the current assessor guide, the draft report goes to your sponsoring director within two weeks of the closing meeting. You have two weeks to respond, and the final report follows within one week of your management responses. It then goes to your next audit and risk committee.
| When | What to do |
|---|---|
| October to November 2026 | Appoint your assessor, agree the twelfth outcome and the terms of reference |
| November to December 2026 | Honest first-pass self-assessment; test the twelve audit outcomes hardest; interim submission point |
| January to March 2027 | Fieldwork for most organisations; fix what you can before the assessor arrives |
| April to May 2027 | Draft and final report, management responses, improvement plan, board briefing |
| By 30 June 2027 | SIRO approval, upload the audit report and improvement plan, publish. Aim for mid-June. |
The full month-by-month plan, including the interim assessment point, is in our complete guide to the CAF-aligned DSPT.
Evidence the assessor will ask for, outcome by outcome
From our experience of DSPT and CAF assessments, this is the evidence that usually decides each mandatory outcome, and where organisations most often fall short. It is a practitioner's view, not NHS England guidance.
A1.b Roles and responsibilities
- What the assessor typically asks for
- Named SIRO, DSPT lead and an owner for each outcome; terms of reference for the group that oversees data security; board or committee minutes showing it is discussed and challenged.
- Where organisations fall short
- Roles exist on paper but nobody has the time or authority to act, or there is no deputy when the lead is away.
A3.a Asset management
- What the assessor typically asks for
- An inventory of the hardware, software, data and services your essential functions depend on, each with an owner, and evidence it is kept current, ideally reconciled against a discovery tool.
- Where organisations fall short
- The inventory is a one-off export that has not been reconciled, and connected medical devices or supplier-hosted systems are missing.
B2.c Privileged user management
- What the assessor typically asks for
- A list of privileged accounts, separate admin accounts from day-to-day ones, MFA on all of them, and records of regular reviews that removed access.
- Where organisations fall short
- Shared or generic admin accounts, and supplier remote-access accounts that nobody reviews.
B2.d Identity and access management
- What the assessor typically asks for
- A joiners, movers and leavers process that works: the assessor will often sample leavers and check their accounts were disabled promptly, including in clinical systems.
- Where organisations fall short
- Network accounts are removed but accounts in clinical and departmental systems stay live.
B3.a Understanding data
- What the assessor typically asks for
- Knowing what data supports your essential functions, where it is held and how it flows, linked to your record of processing activities.
- Where organisations fall short
- A record of processing that exists for GDPR but is not linked to systems, flows or suppliers.
B4.c Secure management
- What the assessor typically asks for
- Systems administered from trusted devices and networks, management interfaces not exposed to the internet, and secure configuration baselines that are actually applied.
- Where organisations fall short
- Administration done from ordinary user laptops, or admin consoles reachable from the internet.
B6.a Culture
- What the assessor typically asks for
- Evidence that leaders promote secure behaviour and that people report incidents and near misses without fear: messages, reporting trends and how reports were handled.
- Where organisations fall short
- Only a policy statement; nothing that shows the culture in practice.
B6.b Training
- What the assessor typically asks for
- Role-based training, with completion data and specialist training for privileged users, developers and the people who respond to incidents.
- Where organisations fall short
- A single annual e-learning module counted as training for everyone, including administrators.
C1.e Personnel skills for monitoring and detection
- What the assessor typically asks for
- Named people or a contracted service with the skills to monitor and investigate, their training, and evidence you can understand and challenge what the provider tells you.
- Where organisations fall short
- Monitoring is outsourced, and nobody inside can explain an alert or hold the provider to account.
D1.c Testing and exercising
- What the assessor typically asks for
- An exercise of your incident response arrangements this year, involving the people who would make decisions, with lessons learned tracked to closure.
- Where organisations fall short
- The exercise happened, but its actions were never closed, or it did not test the plan you actually have.
E4.a Managing records
- What the assessor typically asks for
- A records management policy and retention schedule that are applied to digital records too, with evidence of review and lawful disposal.
- Where organisations fall short
- Retention exists for paper but is not applied to shared drives, email or system data.
Track owners and evidence for every outcome with our free CAF-aligned DSPT evidence tracker.
Choosing your assessor
NHS England encourages organisations to choose providers assured under the NCSC Cyber Resilience Audit scheme, or equivalent, with experience of the CAF. Managing conflicts of interest is your responsibility, and it should be settled before you contract.
The simplest rule: the firm that helped you prepare should not audit the same year. Before you appoint, ask:
- Have you assessed against the CAF-aligned DSPT before, and for organisations like ours?
- Who will do the fieldwork, and what are their qualifications?
- Will you test that controls operate, or only read our documents?
- How will you agree the twelfth outcome with us?
For the DSPT audit itself the auditor has to be able to show independence, so in the same audit year we either help you prepare or audit you, not both. See our DSPT audit and compliance assessment.
Free readiness check: could you face the audit today?
Twelve questions, one for each outcome in a trust, ICB, ALB or CSU audit with the recommended twelfth outcome. Nothing you select is stored or sent; it stays in your browser.
This is an indicative check, not an audit. The real rating depends on your organisation profile's minimum level for each outcome and on what the assessor finds.
Questions we are asked
What is a DSPT audit?
An independent assessment of a defined set of outcomes in your Data Security and Protection Toolkit self-assessment, carried out before you publish. It tests whether the controls work and whether your self-assessment can be relied on.
Which outcomes are audited in 2026-27?
For trusts, ICBs, ALBs, CSUs, OES and genomics organisations: A1.b, A3.a, B2.c, B2.d, B3.a, B4.c, B6.a, B6.b, C1.e, D1.c and E4.a, plus one outcome of your choice. IT suppliers have twelve separate audit areas.
When is the DSPT 2026-27 deadline?
30 June 2027. The audit report and your improvement plan need to be ready before you publish, so most organisations complete fieldwork between January and March.
What is the difference between the risk rating and the confidence rating?
The risk rating reflects how many audited outcomes fall below your minimum expected level. The confidence rating reflects how closely your self-assessment matches what the assessor found.
Can the same firm help us prepare and audit us?
NHS England makes conflicts of interest your responsibility to manage. Our view, and our practice, is that the firm that helped you prepare should not audit you in the same year.
Checked on 30 October 2026 against NHS England's DSPT 2026-27 areas of mandatory audit (1 October 2026) and the guide for CAF-aligned DSPT independent assessors (last updated 9 April 2026). This page is guidance, not legal advice, and UK Cyber-Safe is not part of the NHS.