How to Run a Clinical Audit (PDSA Cycle) That Satisfies Assessors

Assessors do not want to see a number; they want to see that you measured something, changed it, and checked whether the change worked. That loop is a PDSA cycle, and it is what turns a clinical audit into accepted quality-improvement evidence. This guide walks the loop step by step. For the background on how clinical audit fits accreditation, read the clinical audit for RACGP accreditation guide; this is the procedure for running one.

Before you begin

You need a topic worth auditing and a way to measure it. Audits work best on something clinical, common, and changeable in your practice (recall of overdue results, diabetes cycle-of-care completion, allergy fields populated, scripts for a high-risk medication). Decide who will run the cycle and where you will record it, because the record is what becomes CQI evidence at survey time.

Step 1: Pick a topic and a measurable standard

Choose a topic that matters to patient care and set a clear, measurable standard before you look at any data. The standard is the target you are auditing against (for example, "100% of patients on this medication have a documented review in the last 12 months"). A topic without a defined standard produces a number you cannot interpret.

Tie the topic to something real where you can: a known weak spot, a recurring issue, or an area a significant event analysis or a mandatory indicator has flagged. Auditing where you suspect a gap is more useful than auditing where you are confident you already comply.

Step 2: Measure current performance (the baseline)

Pull the data and measure where you actually stand against the standard. Define the patient group, count consistently, and record the baseline figure with the date and method. This is the "before" number, and its credibility depends on you being able to say exactly how you counted it.

Be honest about the baseline even when it is uncomfortable, because the gap between the baseline and the standard is the entire point of the exercise. An audit that starts at 98% and "improves" to 99% rarely convinces an assessor that real improvement happened.

Step 3: Plan and make a change (the PDSA loop)

Decide on one specific change you believe will close the gap, and make it. This is the Plan and Do of PDSA: a new recall workflow, a template change, a staff briefing, a default in the clinical software. Keep the change small and clearly defined, so that when you re-measure you can attribute any movement to it rather than to chance.

Write down what you changed, who did it, and when. A clinical audit that an assessor accepts shows an intervention, not just two measurements with a hopeful gap between them.

Step 4: Re-measure to see if it improved (Study and Act)

After the change has had time to take effect, measure again the same way you measured the baseline. Compare the two figures. If performance improved toward the standard, you have closed or narrowed the gap. If it did not, that is still a valid result: it tells you the change did not work, and the next step is to adjust and run the loop again.

This re-measurement is the step that distinguishes a clinical audit from a one-off data pull. Studying the result and acting on it (standardising the change, or trying a different one) is what makes the cycle genuine quality improvement rather than a snapshot.

Step 5: Document the cycle as CQI evidence

Write the cycle up as a single record: the topic and standard, the baseline, the change you made, the re-measured result, and what you concluded or will do next. This document is the CQI evidence a surveyor reviews, and it is far more persuasive than a spreadsheet of numbers with no narrative.

Keep it with your accreditation evidence and let it feed your next cycle. The same discipline applies to a significant event analysis: both are PDSA loops, and a practice that can show a handful of completed, written-up cycles demonstrates exactly the improvement culture the Standards are looking for.

What good looks like

• The standard was defined and measurable before any data was pulled.
• The baseline is credible, with the patient group and counting method recorded.
• A specific, deliberate change was made between the two measurements.
• Performance was re-measured the same way and compared honestly.
• The whole loop is written up as one CQI record, not a loose spreadsheet.

Common mistakes: measuring without a defined standard, auditing only areas you already pass, collecting a baseline but never making a change, skipping the re-measurement, and leaving the cycle as raw data with no written conclusion.

Frequently asked questions

What is a PDSA cycle in a clinical audit?

PDSA stands for Plan, Do, Study, Act: you plan a change, make it, study the result by re-measuring, then act by standardising what worked or adjusting what did not. A clinical audit run as a PDSA cycle measures a baseline, makes a deliberate change, and re-measures, which is what makes it improvement rather than a one-off count.

How many clinical audits do I need for accreditation?

There is no single magic number; what assessors look for is evidence of an ongoing quality-improvement habit. A small number of completed, properly documented cycles that show measurable change is more convincing than many half-finished audits that never closed the loop.

What makes a clinical audit acceptable to a surveyor?

A defined measurable standard, a credible baseline, a specific change, a re-measurement, and a written conclusion. The surveyor is checking that you closed the loop, so the re-measurement and the documented outcome are the parts that carry the most weight.

How is a clinical audit different from a significant event analysis?

A clinical audit measures performance against a standard across a group of patients; a significant event analysis examines a single event (good or bad) to learn from it. Both are quality-improvement activities structured as PDSA loops, and both count as CQI evidence when documented.

How long does a clinical audit cycle take?

It depends on the topic, but allow enough time between the baseline and the re-measurement for the change to take effect, often several weeks to a few months. Rushing the re-measurement before the change has had time to land produces a result you cannot trust.

Last reviewed 20 June 2026