PICO is a way of breaking a clinical question into parts so that it can be searched and answered: the population you care about, the intervention being considered, the comparison it is being judged against, and the outcome that would count as an answer. It exists because topics are not searchable and questions are. This page is about using it properly, and about the several situations in which it is the wrong frame and a different one fits better.
The four parts, and the one people leave out
Population, intervention, comparison, outcome. The comparison is the part most often skipped, and skipping it is what leaves you with a question nobody can answer: does this work, compared with what? Usual care, a placebo, a different intervention, or nothing at all are four different questions with four different literatures behind them. Fixing the comparison first usually halves the search.
From the four parts to a search
Each element becomes a block of synonyms combined with OR, and the blocks are combined with AND. In practice the population and the intervention blocks carry the work and the outcome block is often left out of the search entirely, because outcomes are reported inconsistently and including them loses studies you want. That is a standard and deliberate move rather than a shortcut, and it is worth stating in the method.
Where PICO stops fitting
It suits questions about whether something works. It fits badly for questions about experience, meaning or process, where a frame built for qualitative work is more useful and the elements are closer to sample, phenomenon of interest and context. Forcing a qualitative question into PICO produces a comparison that does not exist and an outcome nobody measured, and the search that follows finds the wrong literature.
Write it down before you search, and keep it
The question, its elements, and the synonym blocks belong in the record alongside the searches they produced, because six months later you will be asked why a paper is missing and the answer is usually in a synonym nobody thought of. That is what the record this product keeps is for, and the free worksheets on this site cost out the screening the search produces.
A worked example of the four parts
Take a vague topic: does exercise help older people after a fall. As PICO it becomes adults over 65 discharged after a fall (population), a structured exercise program (intervention), usual care or no program (comparison), and further falls within twelve months (outcome). The vague version is unsearchable; the second gives you two searchable blocks and a screening rule. Doing this on paper before touching a database is twenty minutes that saves a day.
Other questions people ask
What does PICO stand for?
Population, intervention, comparison and outcome. Some versions add T for timeframe or S for study design; the core four are what turn a topic into a searchable question.
Do I have to include the outcome in my search?
Usually not. Outcomes are reported inconsistently, so including them as a search block tends to lose relevant studies. Most reviewers search on population and intervention and apply the outcome at screening; say so in the method.
What if my question is not clinical?
Use a frame that fits. PICO is built for whether-it-works questions; qualitative and exploratory questions are better served by frames built around the phenomenon and its context, and forcing PICO onto them invents a comparison that does not exist.
Is PICO only for clinical questions?
It is built for them and fits any whether-it-works question. For questions about experience or process, a frame built around the phenomenon and its context works better, and forcing PICO onto them invents a comparison nobody studied.