PICO Framework: A Question-Formulation Guide for Systematic Review Authors

The PICO framework breaks a research question into four components – Population, Intervention, Comparison, and Outcome – so that a vague topic can be turned into a question that is answerable, searchable in a bibliographic database, and usable as the basis for a systematic review‘s eligibility criteria. It was first described by Richardson et al. in 1995 as part of the evidence-based-medicine literature on formulating a “well-built clinical question,” and the Cochrane Handbook for Systematic Reviews of Interventions identifies it as the recommended starting point for review questions about the effects of an intervention.

This guide is written for systematic-review authors, information specialists, and research librarians building a protocol-driven, PRISMA-reported review – not as a general study-skills introduction. It covers the four core elements and where authors most often get the Comparison element wrong, the full variant family (PICOT, PICOS, PICOTS, PECO, SPIDER, SPICE, ECLIPSE, CIMO, and PCC) and when each is the correct choice instead of PICO, how a PICO question becomes an actual database search strategy, how it maps onto PROSPERO registration and PRISMA reporting, when PICO is the wrong tool entirely, a worked example carried from vague topic to search string, and a fill-in template.

What the PICO Framework Is For

A systematic review cannot be scoped, searched, or registered around a vague topic like “does exercise help diabetes.” PICO forces that topic into four explicit, separately-searchable concepts before any database work begins. Each element becomes: (1) a concept block in the search strategy, (2) an inclusion/exclusion criterion in the review’s eligibility criteria, and (3) a field in the review’s registered protocol. Getting the elements right early is what keeps the search, the eligibility screening, and the final PRISMA flow diagram consistent with each other – a review whose search strategy doesn’t match its registered PICO is a common source of protocol deviation.

The Four Elements, Worked Properly

Element	What it specifies	Where authors go wrong
P – Population	The subjects or problem of interest: condition, demographics, setting-relevant characteristics.	Too broad (“patients”) retrieves everything; too narrow (over-specified demographics) retrieves almost nothing before you’ve even seen what the literature actually reports.
I – Intervention	The treatment, exposure, test, or practice being evaluated.	Bundling two distinct interventions into one question, which then requires two separate search strategies and two separate synthesis plans.
C – Comparison / Comparator	The alternative the intervention is evaluated against.	See below – this is the element most often mishandled.
O – Outcome	The measurable result used to judge effect: a clinical endpoint, a harm, a cost, or a process measure.	Choosing a surrogate outcome when a patient-important or policy-relevant outcome is what the review actually needs to answer.

The comparator/control distinction authors most often get wrong

“Comparator” and “control” are not interchangeable. Comparator is the general PICO element: whatever the intervention is being evaluated against, which could be an active alternative treatment, usual care, a different dose, or no intervention. Control is one specific type of comparator – a no-treatment, placebo, or waitlist arm. Writing “control” into a PICO question when the actual body of literature compares the intervention against an active alternative (a different drug, a different care pathway) narrows the search and excludes studies with genuinely relevant, publishable comparisons. The second common error is treating Comparison as mandatory: for prevalence questions (“how common is X”) and diagnostic-accuracy questions (test performance against a reference standard, not a treatment arm), there is no comparator to force in, and standard methodological guidance is to drop C rather than invent one.

Choosing the Right Framework: PICO’s Variant Family

PICO is the parent format for quantitative intervention questions, but it is not the only question-formulation framework, and using it where it doesn’t fit – a qualitative synthesis, an exposure question with no active intervention, a scoping review – produces a search strategy and eligibility criteria that don’t actually match the evidence being reviewed. This is the single most common real error in framework selection: authors default to PICO because it’s the one they’ve heard of, not because it fits the question.

Framework	Elements	Use it for
PICO	Population, Intervention, Comparison, Outcome	Quantitative intervention/therapy, diagnosis, and prognosis questions – the default for Cochrane-style intervention reviews.
PICOT (sometimes PICOTT)	PICO + Time	When the duration of exposure or follow-up materially changes the answer. See CASRAI’s guide to writing a PICOT question for fully worked conversions.
PICOS	PICO + Study design	Specifying eligible study designs (e.g., RCTs only) directly in the question – common when a PRISMA-P protocol needs the design restriction stated up front rather than left implicit.
PICOTS	PICO + Timing + Setting	Comparative-effectiveness and health-technology-assessment reviews (the AHRQ Evidence-based Practice Centers Methods Guide variant) where care context materially changes the answer. See CASRAI’s PICOTS Criteria entry.
PECO	Population, Exposure, Comparison, Outcome	Observational and environmental-epidemiology questions where there is an exposure rather than an administered intervention – occupational, dietary, or environmental exposure reviews.
SPIDER	Sample, Phenomenon of Interest, Design, Evaluation, Research type	Qualitative and mixed-methods evidence syntheses, where “Population” and “Intervention” don’t map onto a qualitative research question.
SPICE	Setting, Perspective, Intervention/Interest, Comparison, Evaluation	Evaluating a service, project, or policy intervention from a stated perspective – common in health-services and social-care reviews.
ECLIPSE	Expectation, Client group, Location, Impact, Professionals, SErvice	Health-policy and management searches, where the question concerns a service rather than a clinical intervention.
CIMO	Context, Intervention, Mechanism, Outcome	Management and organization-studies reviews, including realist reviews interested in why an intervention works in a given context, not just whether it does.
PCC	Population, Concept, Context	Scoping reviews (the JBI framework) – used instead of PICO when the goal is mapping a body of literature rather than answering a narrow effectiveness question. See CASRAI’s Scoping Review entry and how to conduct a scoping review guide.

How a PICO Question Becomes a Database Search Strategy

Each PICO element becomes a concept block: a set of synonymous or related terms for that concept, combined with Boolean OR, mixing controlled vocabulary (MeSH in MEDLINE, Emtree in Embase, or a database’s own thesaurus) with free-text title/abstract terms to catch records the vocabulary hasn’t indexed yet or indexes inconsistently. The concept blocks are then combined with Boolean AND. In practice:

• Population and Intervention are almost always searched – they’re the two blocks most databases index reliably.
• Comparison is often left out of the search itself and applied during screening instead – searching for it directly risks excluding studies that report the intervention against multiple comparators, only one of which is relevant.
• Outcome is sometimes left out of the search entirely (searched broadly on Population + Intervention, then filtered by outcome at screening), because outcomes are inconsistently reported in titles and abstracts and an overly precise outcome search can silently drop eligible studies.
• Time, when present, is usually applied as a database date-range filter rather than a search term, since duration is rarely indexed metadata.

For the full mechanics of combining concept blocks, nesting parentheses, and using proximity and truncation operators, see CASRAI’s guide to Boolean search operators for literature searching.

From PICO to Eligibility Criteria to PROSPERO and PRISMA

Each PICO element becomes an inclusion or exclusion criterion for study selection: a study is only eligible if it matches the registered Population, delivers the registered Intervention (or a defined range of it), uses an eligible Comparison, and reports an eligible Outcome. This is why the PICO question is finalized before title/abstract screening begins, not derived from whatever the search happens to retrieve.

The same elements are what a PROSPERO registration and a PRISMA-P protocol ask for directly – PROSPERO’s registration form has explicit fields for condition/population, intervention/exposure, comparator, and outcomes, and expects them to match the protocol’s stated eligibility criteria. CASRAI’s systematic review protocol template walks through the full PRISMA-P-aligned document these elements feed into, and the PRISMA and systematic review methodology guide covers how the finalized PICO maps onto the PRISMA 2020 flow diagram and reporting checklist once screening is complete.

When PICO Is the Wrong Tool

• Scoping reviews. A scoping review maps the extent and nature of a body of literature rather than answering a narrow effectiveness question – the JBI methodology uses PCC (Population, Concept, Context) instead. Forcing a PICO structure onto a scoping review’s exploratory question produces an artificially narrow scope. See CASRAI’s literature review vs. systematic review vs. scoping review comparison.
• Purely qualitative syntheses. A qualitative evidence synthesis is interested in experience, meaning, or process, not a measurable effect of an intervention against a comparator – SPIDER, SPICE, or PEO fit these questions; PICO generally does not.
• Methodological reviews. Reviews of measurement properties, diagnostic test accuracy across multiple index tests, or prevalence-only questions often need Comparison dropped or a different framework (e.g., PIRD for diagnostic test accuracy questions with an index and reference test) rather than a standard PICO.
• Reviews asking “why” or “how,” not just “whether.” Realist reviews interested in the mechanisms and contexts that make an intervention work use CIMO, not PICO.

Worked Example: Vague Topic to Search String

Vague starting topic: “Does telehealth work for managing depression?”

PICO elements:

• P: Adults with a diagnosed depressive disorder
• I: Telehealth-delivered psychotherapy (video or phone-based)
• C: In-person, face-to-face delivery of the same or an equivalent psychotherapy
• O: Depressive symptom severity, measured on a validated scale (e.g., PHQ-9)

PICO question: In adults with a diagnosed depressive disorder (P), does telehealth-delivered psychotherapy (I), compared with in-person delivery of the same or an equivalent psychotherapy (C), reduce depressive symptom severity on a validated scale (O)?

Eligibility criteria derived from it: Include studies enrolling adults with a clinician-confirmed depressive disorder; comparing a telehealth-delivered psychotherapy arm against an in-person psychotherapy arm; reporting a validated depressive-symptom outcome measure. Exclude studies with no in-person comparator (e.g., telehealth vs. waitlist only, which is a different question) and studies in pediatric populations.

Boolean search string (concept blocks):

(depress* OR "depressive disorder"[MeSH] OR "major depression") AND (telehealth OR telemedicine OR "video conferencing" OR teletherapy OR "remote delivery") AND (psychotherapy OR "cognitive behavioral therapy" OR counseling) – Comparison (in-person delivery) and Outcome (PHQ-9 or equivalent) are applied at the screening stage rather than searched directly, for the reasons set out above.

This question, its eligibility criteria, and this search strategy are what get entered into the PROSPERO registration and the protocol’s PRISMA-P checklist items before screening begins.

Fill-In PICO Template

Use this to draft a PICO question before building a search strategy or starting a PROSPERO registration:

• Population: ___ (who or what is this about – condition, demographics, setting)
• Intervention: ___ (what treatment, exposure, test, or practice is being evaluated)
• Comparison: ___ (what alternative is it evaluated against – or “none: drop C” if this is a prevalence/diagnostic-accuracy question)
• Outcome: ___ (what measurable result determines the answer – name the specific instrument or endpoint, not just “improvement”)
• Question: In [Population], does [Intervention], compared with [Comparison], affect [Outcome]?

Before finalizing, check the framework fit: if the review is qualitative, use SPIDER or SPICE instead of forcing P/I/C/O labels onto a qualitative question. If it’s a scoping review, use PCC. If there’s an exposure rather than an administered intervention, use PECO. If study design needs to be specified in the question itself, add S for PICOS. If care setting or timing materially changes the answer, use the six-element PICOTS instead of adding those details as free text.

Frequently Asked Questions

What does PICO stand for?

Population (or Patient/Problem), Intervention, Comparison, and Outcome – the four elements used to structure a searchable, answerable research question, most commonly for systematic reviews of intervention effectiveness.

Is the Comparison element always required?

No. Standard methodological guidance is to omit Comparison for prevalence, diagnostic-accuracy, or “meaning” questions where there is no natural comparator – inventing one produces an artificial, unanswerable question.

What’s the difference between PICO and PICOT?

PICOT adds a Time element – the duration of exposure or follow-up needed for the outcome to occur or be measured – to the four-element PICO format. See CASRAI’s guide to writing a PICOT question for worked conversions using the five-element form.

Which framework should I use for a qualitative systematic review?

SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, Research type) or SPICE (Setting, Perspective, Intervention/Interest, Comparison, Evaluation) are built for qualitative and mixed-methods questions; PICO’s Intervention/Comparison structure generally does not map onto them.

Does a PROSPERO registration require a PICO question?

PROSPERO’s registration form asks directly for condition/population, intervention or exposure, comparator, and outcomes – a finalized PICO (or the relevant variant, such as PECO for exposure questions) is what fills those fields. See CASRAI’s PROSPERO Registration entry.

Can I use PICO for a scoping review?

Not directly. Scoping reviews use the JBI PCC framework (Population, Concept, Context) because they map a body of literature rather than answer a narrow effectiveness question. See CASRAI’s guide to conducting a scoping review.

How does a PICO question become a database search strategy?

Each element (typically Population and Intervention, sometimes Outcome) becomes a concept block of controlled-vocabulary and free-text synonyms combined with OR; the blocks are combined with AND. See CASRAI’s guide to Boolean search operators for literature searching for the full mechanics.

Related reading: Systematic Review Abstract Example — A section-by-section walkthrough of the PRISMA 2020 for Abstracts checklist, with a full illustrative composite systematic review abstract example.

Related reading: First Author — First authorship brings career credit, visibility, and CV weight, but usually comes with a disproportionate workload, negotiation costs, and informal exposure if the paper is later questioned.

See also: realist review.