From Clinical Question to Analysis Plan: Designing Research with PICO
Use population, intervention, comparator and outcome elements to create a focused clinical question, hypothesis and analysis target.
What PICO does
PICO separates a clinical question into Population, Intervention or exposure, Comparator and Outcome. A clear PICO supports literature searching, design selection, sample-size planning and definition of the statistical estimand.
Operationalise each element
P: condition, age, setting and key eligibility criteria.
I: measurable treatment, diagnostic strategy or exposure.
C: standard care, placebo, alternative treatment or unexposed group.
O: a valid measure, time point and clinically meaningful interpretation.
“Does a new drug work in diabetes?” is too broad. A focused question specifies the population, add-on treatment, comparator, HbA1c outcome and 24-week horizon.
From question to hypothesis
The clinical question uses clinical language; the statistical hypothesis defines the scale of analysis. If the primary estimand is the adjusted mean HbA1c difference at week 24, the null hypothesis can be expressed on that scale. Non-inferiority and equivalence questions require a prespecified clinical margin.
Common mistakes
Defining the population only by disease name.
Leaving the comparator ambiguous.
Using an unmeasurable outcome such as “improvement”.
Choosing a design that cannot answer the primary question.
Changing the hypothesis after seeing results.
References
Accessed 20 June 2026.