Crossover trial
- Study design
Each patient gets both treatments, one after the other in random order: they are their own control. It suits chronic, stable conditions.
What it is
In a crossover trial each participant receives both interventions in successive periods, with the order assigned at random and a washout period in between. Because each person is compared with themselves, variability between patients is removed and fewer participants are needed than in a parallel trial. It only works in chronic, stable conditions with reversible effects; the main danger is carry-over of the first treatment into the second period.
How it is built
- The order is randomised: half receive A then B; the other half B then A.
- First treatment period, with outcome measurement.
- Washout period without treatment.
- Second period with the other treatment and a new measurement.
- Differences within each patient are analysed (paired analysis).
Worked example
- 24 adults with perennial allergic rhinitis: 12 get an active nasal spray (A) then placebo (B); the other 12 get B then A.
- Each period lasts 4 weeks, with a 2-week washout: total time per patient = 4 + 2 + 4 = 10 weeks.
- The mean symptom score (0 to 10 points) is 4.0 on A and 6.5 on B.
- Mean within-patient difference: 4.0 − 6.5 = −2.5 points in favour of A.
- The effect is checked to be similar in both orders (AB and BA) to rule out carry-over.
- Reading: with only 24 patients, each compared with themselves, A lowered symptoms by 2.5 points.
Crossover versus parallel trial
| Feature | Crossover | Parallel |
|---|---|---|
| Comparison | Within each patient | Between different groups |
| Sample size | Smaller | Larger |
| Suitable conditions | Chronic and stable | Any |
| Own risks | Carry-over and period effects | Imbalance between groups in small samples |
| Time per patient | Longer (two periods and washout) | Shorter |
When to use it
- Used in stable asthma, chronic pain, hypertension, rhinitis and bioequivalence studies in healthy volunteers.
- Drop-outs weigh more: a patient who leaves during the second period loses their comparison.
Common pitfalls
- Using it for diseases that resolve or progress: the second period no longer compares like with like.
- A washout too short for a long-acting treatment: carry-over contaminates the second period.
- Analysing the data as two independent groups instead of a paired analysis.
Test yourself
Think of the answer, then tap the question to see it.
What is the main advantage of a crossover trial?
Each patient is their own control, which reduces variability and allows fewer participants.
What is carry-over?
The persistence of the first treatment's effect into the second period, which biases the comparison.
Why is a crossover trial unsuitable for an acute infection that resolves?
Because after the first period the patient no longer has the disease, so there is nothing to compare in the second.
Connections
Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.
Sources: Chapter 23: Including variants on randomized trials (Cochrane, 2024); Chapter 8: Assessing risk of bias in a randomized trial (Cochrane, 2024).
Spotted a mistake?We are always improving for the community. Tell us what is wrong and our team will review it.Report a mistake
Representative model, for educational purposes only.