Communicating risk, Cates plots, decision aids, audit and prescribing data
The applied end of the RCGP data-interpretation list: Cates plots and decision aids for shared decisions, the audit cycle, control charts, and prescribing measures such as ADQ and STAR-PU.
Reviewed and kept current
Last editorial review: 28 September 2026 · Next scheduled refresh: March 2027.
Clinical vignette
Begin with the patient. Read the case, weigh the options, then work through the unit.
Mr M is considering a preventive treatment to reduce his risk of stroke. You show him a Cates plot of 100 people like him, treated for ten years.
Without treatment, 85 have no stroke and 15 have a stroke. With treatment, 85 faces are green (no stroke anyway), 5 are yellow (stroke prevented by treatment) and 10 are red (stroke despite treatment).
Mr M asks: “So if I take it, I won’t have a stroke?” He wants a straight answer. He also wants to know whether it is worth taking every day for ten years.
Options to weigh
- Quote the relative risk reduction of 33%.
- Tell him that the treatment will prevent his stroke.
- Explain that only the 5 yellow faces benefit, which gives a number needed to treat of 20 over ten years.
- Use both framings: 85 would have had no stroke anyway, 5 avoid a stroke, and 10 still have one.
Reflect
How would you use this plot so that Mr M can decide, and what would you be careful not to say?
What this unit covers
Pocket-sized and self-contained: three AKT-style questions, the answers explained, an expanded discussion with hyperlinked guidance, a decision algorithm and a certificate.
By the end of this unit you will be able to
Five steps, in order
Answer first, then learn. Each section unlocks as you progress, and your progress is saved in this browser.
1Three single best answer questions
Commit to an answer for all three single best answer questions, then submit. Sections 2–4 open on submission; your certificate opens at 2/3 or better.
Question 1 of 3
A Cates plot shows 100 faces for people like your patient treated for ten years. Without treatment, 85 have no stroke and 15 have a stroke. With treatment, the chart shows 85 green faces (no stroke anyway), 5 yellow faces (stroke prevented by treatment) and 10 red faces (stroke despite treatment).
What is the number needed to treat for ten years to prevent one stroke?
Question 2 of 3
A practice audits whether patients on methotrexate have had blood monitoring at the recommended interval. The standard is 95%; 71% compliance is found. The practice introduces a recall system.
Which step completes the audit cycle?
Question 3 of 3
An ICB wants to compare antibacterial prescribing fairly between practices whose list sizes and age–sex profiles differ considerably.
Which denominator is most appropriate?
2Answers explained
Question 1
Correct answer D. 20
The yellow faces are the people who benefit: 5 in 100 = ARR 5%.
Cates plots display absolute benefit visually. This makes them ideal for shared decision-making and matches NICE advice to use absolute numbers with a common denominator.
Question 2
Correct answer B. Re-measuring compliance against the same standard after the change
Audit compares practice against an explicit standard, implements change, then re-audits against the same criteria to see whether care has improved — “closing the loop”. Without re-measurement, it is only data collection.
Audit asks “Are we doing what we should?”; research asks “What should we be doing?” (generalisable new knowledge, needs ethics approval); service evaluation asks “What standard does this service achieve?” without a pre-set standard.
Question 3
Correct answer E. Specific therapeutic group age–sex related prescribing unit (STAR-PU)
STAR-PUs weight each practice’s list by age and sex according to the expected need for a specific therapeutic group (e.g. antibacterials), so that practices with more elderly patients or more children are compared fairly (items per STAR-PU).
- ADQ measures volume (the typical daily maintenance dose in England), used as a numerator (e.g. ADQs per STAR-PU).
- Items and cost alone ignore population differences.
3Expanded discussion, worked examples & guidance
1. Communicating risk well
People understand absolute risks as natural frequencies (“5 out of 100”) far better than relative risks or percentages. Good practice (NICE NG197 on shared decision-making):
- use a consistent denominator (per 100 or per 1,000) and a stated time frame;
- present both benefits and harms in the same format;
- avoid relying on relative risk reductions alone;
- use visual aids — icon arrays such as Cates plots — and positive and negative framing (“95 won’t have a stroke” and “5 will”);
- use a patient decision aid where one exists, e.g. NICE’s decision aid on statins for preventing cardiovascular disease.
Worked example 1 — building a Cates plot from a trial
Ten-year CVD risk 20% without statin; atorvastatin 20 mg reduces relative risk by about 30%.
- Treated risk ≈ 20% × 0.7 = 14%.
- Out of 100: 80 green (no event anyway), 6 yellow (event prevented), 14 red (event despite treatment).
- ARR 6% → NNT ≈ 17 over 10 years.
Answer. “If 100 people like you take a statin for ten years, about 6 avoid a heart attack or stroke; 80 would have been fine anyway; 14 still have one.” (Illustrative figures; use QRISK3 and the NICE decision aid in practice.)
2. Audit, research, service evaluation and QI
| Question | Standard? | Ethics approval? | |
|---|---|---|---|
| Research | What should we do? (new, generalisable knowledge) | Hypothesis | Yes (REC) |
| Clinical audit | Are we doing what we should? | Yes — explicit criteria and target | No |
| Service evaluation | What does this service achieve? | No | No |
| Quality improvement (PDSA) | How can we improve, in small rapid cycles? | Aim and measures | No |
Audit cycle: choose topic → set standard (criteria + target, e.g. from NICE) → measure → compare → implement change → re-audit. Significant event analysis examines a single event in depth (what happened, why, what was learnt, what changed).
3. Comparing practices: control charts and funnel plots
Plotting each practice’s rate against its list size, with 95% and 99.8% control limits, forms a funnel plot. Small practices vary more by chance. A practice outside the 99.8% limit is unlikely to differ by chance alone, and warrants a supportive look at case-mix, coding and care. League tables without CIs mislead: the “worst” practice is often just the smallest. Statistical process control (SPC) charts plot a measure over time with control limits; points outside the limits, or runs of points on one side of the mean, indicate special-cause variation.
4. Prescribing data
| Measure | What it counts | Use |
|---|---|---|
| Items | Number of prescription items dispensed | Simple volume; ignores quantity and population |
| Net ingredient cost (NIC) | Basic drug cost | Spend |
| DDD (defined daily dose, WHO) | Assumed average adult maintenance dose | International volume comparisons |
| ADQ (average daily quantity) | Typical daily dose in England | Volume numerator for NHS comparisons |
| ASTRO-PU / STAR-PU | List weighted by age and sex (overall, or for a therapeutic group) | Fair denominator between practices |
Worked example 2
Practice A: 3,000 antibacterial items; 3,500 antibacterial STAR-PUs. Practice B: 2,400 items; 2,000 STAR-PUs.
- A: 3,000/3,500 = 0.86 items per STAR-PU.
- B: 2,400/2,000 = 1.20 items per STAR-PU.
Answer. Practice B prescribes more relative to its population need, despite issuing fewer items.
- Audit needs a standard; without one it is a service evaluation.
- Regression to the mean: practices selected for being extreme tend to look better next time even without intervention.
- In a Cates plot, only the yellow (benefit) faces represent the effect of treatment.
References and further reading
Guidance checked at the last editorial review (28 September 2026). Always use the current version.
- NICE NG197 — Shared decision making
- NICE — Patient decision aid: statins for preventing cardiovascular disease (NG238)
- Cates plots — Visual Rx / NNT resources
- NHS Business Services Authority — prescribing measures (ePACT2, ADQ, STAR-PU)
- Healthcare Quality Improvement Partnership — Best practice in clinical audit
- RCGP — Data interpretation topics in the AKT (PDF)
4Decision algorithm
Supporting a shared decision about a preventive treatment
Start
Establish the patient’s baseline risk (e.g. QRISK3, FRAX, CHA₂DS₂-VASc) and their values.
Action
Find the relative effect of treatment (RRR) from the best evidence (NICE, systematic review).
Action
Convert to absolute terms: ARR = baseline risk × RRR; NNT = 1/ARR over the stated time.
Action
Do the same for harms: ARI and NNH.
Action
Show a Cates plot or decision aid: “Out of 100 people like you over 10 years…”
Decision
Does the patient judge that the benefits outweigh the harms and burdens?
Outcome
Record the discussion, the numbers used and the decision (NICE NG197).
Revision summary; apply clinical judgement and local pathways.
5Certificate of completion
You have met the pass mark. Enter your details to issue your personalised certificate of completion, then save it as a PDF for your appraisal or revalidation portfolio.
Your certificate is issued immediately as a PDF. Your details stay in your browser and are not sent to MD Acumen.
Your completion and score have also been recorded by your organisation’s learning platform.
MRCGP AKT · CAS Programmes · For institutions · Legal & regulatory · enquiry@mdacumen.com
Educational content for healthcare professionals; apply clinical judgement and current local guidance. MRCGP is a registered trademark of the Royal College of General Practitioners. These units are independent, curriculum-aligned revision resources and are not endorsed by the RCGP.
© 2026 MD Acumen Ltd · Registered in England and Wales · Company no. 16538952 · ICO registration ZC105897