Writing a Clinical Audit Report
The audit cycle in writing, standards, measurement against them, and the changes that close the loop. Structure for a report that drives improvement.
Clinical audit measures practice against a defined standard and drives improvement. It is not research, it does not generate new knowledge, but a distinct quality-improvement activity, and writing it up follows the audit cycle rather than the research paper format.
The Audit Cycle
Audit is a loop: set a standard, measure current practice against it, identify gaps, implement change, then re-audit to confirm improvement. A report that stops at "we measured and found gaps" has done half the job. Closing the loop with re-audit is what makes it audit rather than a survey.
Report Structure
- Title and aim: what practice is being audited and why.
- Background: the clinical context and why this matters for patient care.
- Standards and criteria: the explicit, measurable benchmark, ideally from national guidelines (e.g. NICE), with a target level (e.g. "100% of patients should have X documented").
- Methodology: what data, over what period, from where, and how sampled.
- Results: current performance against each standard, with figures and tables.
- Discussion: where and why practice fell short.
- Recommendations and action plan: specific, assigned, time-bound changes.
- Re-audit plan: when the loop will be closed.
Standards Must Be Explicit and Sourced
The whole audit hinges on the standard. It should be measurable and drawn from authoritative guidance, not invented. "Patients should be treated well" is not auditable, whereas "All patients with suspected sepsis should receive antibiotics within one hour (per the Sepsis Six)" is.
Results and Recommendations
Report performance as a clear percentage against target, using charts where they aid comprehension. Then make recommendations that are actionable and owned: not "improve documentation" but "introduce a checklist sticker. Ward pharmacist to audit weekly. Re-audit in three months." Recommendations that no one owns never happen.
Audit vs Research, Keep the Distinction
Audit compares practice to an existing standard. Research asks what the standard should be. This matters for ethics: audit usually does not require research ethics approval, but you must confirm your institution's governance rules.
Checklist
- Is the standard explicit, measurable and sourced?
- Are results reported against target clearly?
- Are recommendations specific, owned and time-bound?
- Is a re-audit planned to close the loop?
- Have I confirmed with governance that this is audit and not research?
- Is the sample and time period stated precisely enough to repeat?
Audit, Research or Service Evaluation?
This is the first question, and getting it wrong has consequences. Audit measures practice against an existing standard. Research generates new knowledge and needs ethics approval. Service evaluation describes current practice without judging it against a standard.
| Clinical audit | Service evaluation | Research | |
|---|---|---|---|
| Question | Does practice meet the standard? | What does this service currently do? | What should practice be? |
| Standard | Pre-existing and explicit | None applied | Being established |
| Intervention | None outside normal care | None | May allocate or intervene |
| Ethics approval | Usually not required | Usually not required | Required |
| Governance | Local registration usually required | Local registration usually required | Formal approval route |
The Health Research Authority publishes a decision tool for exactly this question, and the sensible move is to run your project through it and keep the output. If your project randomises anything, withholds standard care, or is designed to produce generalisable new knowledge, it is research however you label it.
The Audit Cycle
- Select a topic where practice matters, evidence exists and change is achievable.
- Define the standard and target, drawn from authoritative guidance, with an explicit percentage.
- Collect data against a pre-specified sample and period.
- Compare with the standard and identify where and why practice falls short.
- Implement change, with named owners and timescales.
- Re-audit to establish whether the change worked.
An audit that stops at step four is not an audit, it is a survey of practice. The loop closes at step six, and reports that never reach it are the commonest weakness in student and trainee audit work.
Writing a Standard That Can Be Measured
The standard carries the whole project. It needs a source, a criterion and a target.
| Weak | Auditable |
|---|---|
| Pain should be managed appropriately. | 100 per cent of patients presenting with suspected fracture should have a pain score documented within 30 minutes of arrival (source: local trust guideline, 2025). |
| Documentation should be adequate. | 95 per cent of discharge summaries should record allergy status (source: national standard). |
| Follow-up should be timely. | 90 per cent of patients should receive a follow-up appointment within six weeks of discharge. |
Note that each auditable version names a source. A standard you invented is not a standard, and comparing practice against your own opinion is the error that most often makes an otherwise competent audit unusable.
On targets, 100 per cent is right for absolute safety requirements and unrealistic for most else. Where the guidance specifies a figure, use it. Where it does not, set one, justify it, and say who agreed it.
Presenting the Results
Report performance against target, not raw numbers alone. A table with the criterion, the target, the achieved percentage, the numerator and denominator, and a pass or fail column lets a reader see the position instantly.
Give absolute numbers alongside percentages. "60 per cent compliance" from a sample of five patients is three cases, and a reader who sees only the percentage will overestimate what you have shown. Small samples are common in trainee audit and are acceptable, provided the limitation is visible.
Recommendations That Change Something
| Weak | Actionable |
|---|---|
| Staff should be reminded of the guideline. | The clinical lead will add allergy status as a mandatory field in the discharge template by 1 October, removing reliance on recall. |
| Training is needed. | The 20-minute induction session will include the pain-score standard from the next intake, with attendance recorded. |
| Documentation should improve. | A prompt will be added to the triage proforma, and compliance re-audited after three months. |
The strongest recommendations change the system rather than exhorting individuals. A mandatory field, a proforma prompt or a checklist works when everyone is busy. A reminder to be careful does not survive a bad shift, which is precisely when compliance matters.
Why Audits Fail to Change Anything
Most trainee audits are competently measured and change nothing, and the reason is usually predictable. The project is timed to a rotation rather than to the problem, so the person who found the gap has left before the re-audit. The recommendation asks busy people to remember something. Or the findings are presented once at a meeting and never reach the people whose practice they describe.
Three things help. Hand the project over formally, naming who will run the re-audit and when. Change the system rather than the intention, since a mandatory field outlasts any individual's diligence. And present the results to the team whose practice was audited, not only to the audit department, because the people who can change practice are the ones who need to see the number.
Common Mistakes and Fixes
| Mistake | Fix |
|---|---|
| Standard invented rather than sourced | Cite national or local guidance |
| No re-audit | Plan and date it. The cycle is the method |
| Percentages without denominators | Report both |
| Recommendations with no owner | Name who does what by when |
| Project is really research | Run the HRA decision tool before starting |
| Patient-identifiable data in the report | Anonymise at collection, not at write-up |
| Not registered with the audit department | Register first. Most trusts require it |
Where Ethical Support Fits
Clinical audit departments exist to help with exactly this, from registration and proforma design through to data collection methods. Statistical advice is legitimate. So is asking a supervising clinician whether your chosen standard is the right one.
What stays yours is the data collection, the honest reporting of the result and the recommendations. An audit that quietly reports better compliance than it found is a governance problem rather than an academic one.
Frequently Asked Questions
Does clinical audit need ethics approval?
Usually not, but it normally needs local registration and information governance sign-off. Confirm with your audit department before collecting anything.
How many cases should I include?
Enough to be meaningful for the standard and the setting. Consecutive cases over a defined period are preferable to a convenience selection, since selection bias is otherwise hard to exclude.
What if compliance is already 100 per cent?
That is a valid result. Report it, consider whether the standard was set too low, and say what maintains the performance.
Can I audit my own practice?
Yes, and it is common. Acknowledge the potential for bias in data collection, and where possible have someone else extract a sample to check.
How soon should the re-audit happen?
Long enough for the change to take effect, commonly three to six months. Say when and who will do it, particularly if you will have rotated by then.
Can audit be published?
Yes, in quality improvement journals. Reporting guidelines exist for quality improvement work, and the EQUATOR Network indexes them.
Your Next Step Today
Write your standard as a single sentence containing a criterion, a target percentage and a named source. If you cannot name the source, you do not yet have a standard, and everything downstream will be unusable until you do.
Trusted Sources
- Healthcare Quality Improvement Partnership, best practice guidance on clinical audit. Accessed 11 August 2026.
- Health Research Authority, Research planning, including guidance on whether a project is research. Accessed 11 August 2026.
- Health Research Authority. Accessed 11 August 2026.
- EQUATOR Network, reporting guidelines including those for quality improvement. Accessed 11 August 2026.
This is general study and writing guidance, not clinical or governance advice. Registration and approval requirements are set locally, so follow your own organisation's rules.
Related services
Related guides
- How to Write a Medical Case Report (CARE Guidelines)
- Systematic Review and Meta-Analysis with PRISMA
- Report Writing vs Essay Writing: Structures Compared
- Medical Research: Data Collection, Analysis and Reporting
All academic writing guides or browse our academic writing services.