Clinical audit vs research vs quality improvement in nursing is a distinction you should resolve before choosing a project method. These activities can use similar data, but their purposes differ. Classifying the work correctly helps you align the question, design, local governance and dissertation claims.
This UK-focused guide explains the differences through practical examples and a planning checklist. It does not determine approval for an individual project. Use the relevant university and organisation processes to obtain a documented classification before collecting data or changing care.
Clinical audit vs research vs quality improvement in nursing: quick answer
Clinical audit examines care against an explicit standard and supports action to improve compliance. Research seeks to generate new knowledge through a defined research question and suitable design. Quality improvement develops and tests changes to improve a service or process in its context.
These activities can interact. An audit may identify a gap that leads to improvement work, and a research study may evaluate an improvement intervention. The correct classification depends on the actual purpose and procedures, rather than the label written on the proposal.
The Health Research Authority’s decision tool helps identify whether a study meets the UK definition of research. Read its questions carefully and discuss the outcome with the relevant local office (HRA, n.d.).
Compare the main purpose of each activity
| Activity | Main question | Typical starting point | Useful result |
|---|---|---|---|
| Clinical audit | Does current care meet a defined standard? | An explicit criterion and target | A compliance gap and an action plan |
| Research | What new knowledge can this investigation generate? | A research question and justified design | Findings addressing that question with stated limits |
| Quality improvement | How can this local system perform better? | A service problem and measurable aim | Learning about changes and their consequences |
| Service evaluation | How is this existing service performing? | A question about current provision | A contextual assessment to inform decisions |
The table is a planning aid, rather than a substitute for formal determination. A project can contain more than one component, and seemingly small design changes can alter the classification.
For example, using routine records does not automatically make work an audit. A record-based study could address a research question. Likewise, collecting interviews does not automatically make a project research; the purpose and procedures require assessment.
What distinguishes a clinical audit?
An audit requires a clear standard against which practice can be compared. The criterion describes what should happen, and the target specifies the intended level of performance. Identify a defensible source, such as an applicable guideline or agreed policy.
“Patients should receive good communication” is too broad to audit reliably. A usable criterion specifies the required communication action, eligible cases and accepted evidence. The definition must be precise enough for consistent review.
Audit is more than collecting a compliance percentage. The findings should inform action, and subsequent review can assess whether practice has changed. Reporting a gap without an improvement response leaves the practical cycle incomplete.
The HQIP guide provides a source for understanding audit alongside research and service review. Use it to support the classification discussion and your explanation of the chosen approach (HQIP, 2011).
Invented audit example
Suppose an organisation’s approved discharge policy requires eligible patients to receive a documented follow-up contact route. A student reviews a defined sample of records to assess whether that criterion is met.
The report would identify the policy source, eligible population, collection period and compliance result. It would also describe proposed actions and a planned repeat review. It should not claim to have tested every possible method of discharge education.
This example depends on an actual applicable standard in a real project. Do not invent a policy requirement or target to make a dissertation look like an audit.
What distinguishes research?
Research begins with a question intended to generate new knowledge. The design must provide a defensible way to answer it. Depending on the question, this might involve quantitative, qualitative or mixed methods.
For example, a study could investigate how patients experience nurse-led discharge communication across defined settings. Another could compare an intervention under a research design intended to estimate its effect.
The method is not defined solely by whether statistics are used. A qualitative interview study can be research, and an audit can involve numerical analysis. The purpose and design are more informative than the presence of a p-value.
Use our nursing methodology decision guide to align the question with a design. Check that recruitment, measurement and analysis can address the intended contribution.
Invented research example
Imagine a proposed study examining patients’ accounts of uncertainty after discharge. Its question concerns experiences and meanings, with a justified qualitative sampling and analysis approach.
The study might inform communication practice, but that practical relevance does not automatically make it a local QI project. Its research purpose, procedures and planned knowledge contribution need to be stated clearly.
The findings would describe the analysis and limitations, including whose experiences were represented. They would not establish the effectiveness of an intervention that was never implemented or compared.
What distinguishes quality improvement?
QI starts with a problem in a service and a plan to improve its performance. It commonly involves testing changes, studying what happens and adapting the approach through local learning.
The project needs a measurable aim and evidence that the problem exists. It also needs a plausible connection between the proposed change and the desired result. A new form or teaching session is an activity, rather than proof of improvement.
For the wider project workflow, see our nursing QI guide. The present article concentrates on choosing the project type and avoiding misclassification.
QI does not become research merely because the report is carefully written. Conversely, calling a comparative investigation “QI” does not exempt it from research assessment when its actual design meets that definition.
Invented improvement example
A ward team identifies inconsistent communication of the follow-up contact route. It introduces a short prompt, tests it during a small number of consultations and adapts the workflow after feedback.
The team tracks patient understanding, prompt use and a plausible balancing consequence, such as consultation time. The report describes the setting, changes, measures and patterns over time.
This differs from simply checking compliance with a policy. It studies how a change works in the local system, while still requiring appropriate governance and careful claims about causation.
Where does service evaluation fit?
Service evaluation examines an existing service to inform decisions about its provision. It may ask how users experience the service, whether access is timely or whether the current model meets local needs.
For instance, a clinic may evaluate patient feedback on its existing telephone follow-up pathway. The purpose might be to inform a local decision about the service, rather than to compare compliance against a fixed standard.
Do not use “service evaluation” as a convenient label for work whose actual purpose is different. Describe the question, data and planned use, then obtain the appropriate local determination.
Some projects combine evaluation with improvement. Keep the components explicit so that readers can distinguish assessment of the current service from testing a new change.
The same data can support different questions
Consider a set of discharge records. An audit might compare their documentation with a policy criterion. A service evaluation might describe how the current pathway operates. A research study might examine a relationship between documented features and a defined outcome.
The data source alone does not settle the classification. Nor does the fact that information was collected routinely. Reuse still requires a justified purpose, appropriate access and a method suited to the question.
A useful planning exercise is to write the intended question in one sentence without naming a project type. Then describe what you will actually do. The classification should follow those details, rather than the reverse.
If the procedures change, revisit the determination. Adding a new intervention, recruitment approach or comparison can materially change the project. A previous classification may no longer describe the revised work.
Resolve classification before approvals and collection
Begin with your supervisor and the organisation’s relevant research, audit or improvement office. Provide a concise protocol describing the purpose, population, setting, procedures and data handling. Request a written record of the decision.
The HRA explains that local audit and service-evaluation processes are organisation specific. Work classified as non-research can still need local review and university approval. “Not research” should not be written as “no governance required” (HRA, 2022).
Research pathways also depend on the study and jurisdiction. Check the current requirements with the responsible office rather than assuming that one approval applies to every UK setting.
Do not start collecting data while waiting for a determination that is required for the work. Build realistic time for classification, permissions and any revision into the project schedule.
Check student eligibility and supervision
A nursing dissertation must fit both the academic brief and the setting’s requirements. A project acceptable to a service may not meet the university’s assessment requirements, and a suitable academic idea may be impractical in the placement.
The HRA’s student research guidance should be checked when planning relevant NHS or health and social care research. Discuss the applicable pathway with your supervisor rather than assuming that every student can undertake every design.
Keep the project feasible. Access to a workplace does not automatically authorise recruitment, record review or a change in care. Identify who can approve each part before committing to the proposal.
Consider ethical issues in every project type
Ethical thinking applies to audit, evaluation and improvement as well as research. Consider privacy, burden, fairness and the possibility of unintended consequences. A non-research classification does not make these issues disappear.
If interviewing patients, assess whether the invitation could feel connected to their care. If involving colleagues, consider power relationships and whether participation appears voluntary. Use the locally approved approach.
Only collect information needed for the agreed purpose. Plan secure storage, access and reporting. Small groups can be identifiable even when names have been removed, especially in a single ward or specialist service.
Describe these decisions in the methodology. Avoid vague statements such as “all ethical issues were addressed” without explaining the actual review and safeguards.
Use a classification decision record
| Planning question | Evidence to retain |
|---|---|
| What is the primary purpose? | A concise question or measurable aim |
| Is there an explicit standard? | The source and applicable criterion |
| What procedures will be introduced? | A description of care, measurement and recruitment |
| What knowledge or decision is intended? | The planned use and limits of findings |
| Who determined the classification? | The relevant office or authorised reviewer |
| What permissions are required? | A documented list and status |
| What changes would trigger review? | A revision and amendment process |
This record makes the reasoning visible. It also helps your supervisor identify a mismatch before the project develops further. Keep the version associated with the approved protocol.
Use our proposal feasibility guide to connect the classification with practical access, timescale and method decisions. A convincing proposal explains both what the work intends and how it can proceed.
Can an audit lead to a QI project?
Yes. An audit can identify a gap against a standard, and a team can then test changes intended to close that gap. The repeat audit may form part of the subsequent evaluation.
State the relationship clearly. The initial audit establishes the recorded baseline against the criterion. The improvement component describes the changes, implementation and learning. The repeat review assesses what happened after action.
Do not retrospectively claim a complete improvement cycle if no change or repeat assessment occurred. A dissertation should report the work actually undertaken, including unfinished stages and practical limitations.
Likewise, a research study can evaluate a QI intervention. The intervention’s improvement purpose does not by itself determine the classification of the study evaluating it.
Does intending to publish make the project research?
Publication is not a reliable shortcut for classification. Audit, evaluation and improvement work can be reported and shared. The underlying purpose and design still need assessment.
A journal’s reporting requirements are also separate from local permission to conduct the work. A well-prepared manuscript cannot retrospectively replace missing authorisation.
When writing for publication, use a reporting guideline suitable for the project. SQUIRE 2.0 is relevant to reporting systematic healthcare improvement efforts and helps authors explain the intervention and context (Ogrinc et al., 2016).
Do not force every dissertation into the same reporting checklist. A qualitative research study and a clinical audit need different methodological explanations, even if both concern discharge communication.
Explain the classification in your dissertation
Your methodology should state what type of work was undertaken and why. Connect the explanation to the purpose and procedures, then identify the local determination and relevant approvals accurately.
For an audit, explain the standard, criteria, target and sampling. For research, justify the design and analytical approach. For QI, describe the local aim, change process, measures and evaluation.
The discussion should respect those boundaries. An audit percentage does not establish a universal causal relationship. A short local improvement project does not automatically demonstrate effectiveness across all services.
Use precise wording about review. If a project was classified as non-research and registered locally, say that. Do not describe the classification letter as an NHS research ethics approval if that is not what it was.
Common classification mistakes to avoid
- Choosing the easiest approval route before defining the purpose.
- Assuming that a questionnaire always means research.
- Calling every routine-record review an audit.
- Using “QI” to avoid assessing a research design.
- Inventing an audit standard or compliance target.
- Equating non-research with no review or permission.
- Treating publication plans as the deciding criterion.
- Reporting a completed cycle when key stages never occurred.
If your proposal contains several of these uncertainties, simplify it before collection. A narrower, correctly classified project is easier to execute and explain than an ambitious design with unresolved governance.
Help aligning the project with your brief
Our nursing methodology support can help you clarify the purpose, structure the protocol and prepare questions for your supervisor. Formal classification and authorisation remain with the relevant university and organisation.
Frequently asked questions
What is the main difference between audit and research?
Audit compares care with an explicit standard. Research addresses a question intended to generate new knowledge. The actual procedures and purpose should be assessed together.
Is every QI project non-research?
No. An improvement intervention can be investigated through a research design. Obtain a determination based on the planned work rather than its label.
Does a non-research project need any approval?
It can still require local review, registration, permission and university approval. Check the processes applicable to your organisation and project.
Can service evaluation include interviews?
It may, but the purpose, procedures and ethical issues need assessment. The data-collection method alone does not establish the classification.
Can an audit or evaluation be published?
These activities can be shared and published when appropriate. Publication plans do not replace project classification or the permissions needed to conduct the work.
Start with purpose and obtain a documented decision
Choosing between audit, research and QI begins with the question and intended use of the work. Describe the procedures honestly, involve the appropriate reviewers and retain the determination. This gives your nursing dissertation a clearer method and more defensible conclusions.
References
- Health Research Authority (n.d.) Is my study research? Available at: HRA decision tool (Accessed: 30 September 2026).
- Health Research Authority (2022) Question and Answers: Student Eligibility Criteria. Available at: HRA guidance (Accessed: 30 September 2026).
- Healthcare Quality Improvement Partnership (2011) HQIP guide for clinical audit, research and service review. Available at: HQIP guide (Accessed: 30 September 2026).
- Ogrinc, G. et al. (2016) ‘SQUIRE 2.0: revised publication guidelines from a detailed consensus process’, BMJ Quality & Safety, 25(12), pp. 986–992. Available at: Full article.
