Question frameworks for systematic reviews are structured templates that break a review question into named concept blocks, and the three you will meet most often are PICO for intervention questions, PCC for scoping and prevalence questions, and SPIDER for qualitative and mixed-methods questions. The right framework depends entirely on the kind of question you are asking, not on personal preference.
Why the framework choice is a methodological decision
Picking a framework is not box-ticking. Each one assumes a different shape of evidence, so choosing the wrong one forces your question into a mould that does not fit and quietly distorts the search. An intervention question squeezed into a qualitative framework will miss trials; a qualitative question forced into PICO will demand a comparator that does not exist. The framework you choose also dictates which effect measures or synthesis approach is even possible, so it belongs in your protocol from the start.
The three frameworks compared
PICO: intervention and effectiveness
PICO splits a question into Population, Intervention, Comparator, and Outcome, and it is the default for questions about whether a treatment works. Adding study design gives PICOS. It maps cleanly onto a randomised-trial evidence base and onto a meta-analysis. We unpack each element in the PICO framework.
PCC: scoping and mapping
PCC stands for Population, Concept, and Context, and it deliberately drops the comparator and outcome. That looser structure suits broad mapping questions where you want to chart what exists rather than test an effect, which is exactly the territory of a scoping review. Because it is broad, a PCC question usually returns more records, so plan screening capacity accordingly.
SPIDER: qualitative and experience
SPIDER covers Sample, Phenomenon of Interest, Design, Evaluation, and Research type. It was built for questions about lived experience and perception, where a population-and-intervention split makes little sense. SPIDER searches tend to favour sensitivity less than PICO, so they pair well with broader grey literature searching and citation chasing to catch studies that databases index poorly.
The same question framed three ways
The clearest way to feel the difference is to watch one broad topic, smartphone applications and physical activity, take a different shape inside each framework. A PICO version reads: “In sedentary adults (Population), do smartphone activity applications (Intervention) compared with no application (Comparator) increase daily step count (Outcome)?” That phrasing presumes a measurable effect and a control group, so it points straight at randomised trials and a poolable forest plot.
A PCC version of the same topic deliberately removes the effect: “What is known about how smartphone activity applications (Concept) are used by adults (Population) in community settings (Context)?” That maps the field, including study types, populations studied, and gaps, without asking whether anything “works”. A SPIDER version turns instead to meaning: “How do adults (Sample) experience using activity applications (Phenomenon of Interest) in qualitative interview studies (Design and Research type)?” Three legitimate reviews, three incompatible searches, one topic. The framework is what decides which of the three you are actually doing.
Lesser-known frameworks worth knowing
The three headline templates do not exhaust the field, and forcing an awkward fit is worse than reaching for a sharper tool. For questions about harm or risk factors, PECO swaps Intervention for Exposure, which reads more honestly when no one deliberately assigned the exposure. For prevalence and incidence work, CoCoPop (Condition, Context, Population) names the three elements an epidemiological count actually needs. Diagnostic accuracy reviews lean on the index test, reference standard, and target condition triad rather than an intervention and a comparator. And policy or qualitative questions sometimes suit SPICE (Setting, Perspective, Intervention, Comparison, Evaluation). The lesson is not to memorise every acronym but to recognise when the standard three do not capture your elements cleanly.
Whichever you adopt, the framework decides whether a numeric synthesis is even on the table. A PICO or PECO question with a comparator can produce a pooled estimate, the kind you can preview with a meta-analysis calculator once the studies are in; a PCC mapping question almost never does, because there is no shared effect to combine. Reviewers who pick the template late, after the search has run, often discover they have collected studies that cannot answer the question they meant to ask.
Common mistakes when choosing a framework
Most framework errors are visible long before the search, if you know what to look for:
- Forcing a comparator that does not exist. Qualitative and prevalence questions have no control arm, so demanding a PICO “C” either invents a false contrast or quietly narrows the review to a sliver of the literature.
- Treating the framework as cosmetic. Filling in the letters after writing the question backwards is box-ticking. The template should generate the question, then seed the eligibility rules and the search concept blocks.
- Over-specifying a mapping question. Bolting a precise outcome onto a PCC question collapses a scoping review into a narrow effectiveness review, defeating the point of mapping.
- Mixing whole frameworks in one question. Borrowing one element is fine; stitching PICO and SPIDER together usually means the underlying question is really two questions that need separating.
Each of these is far cheaper to catch at the protocol stage than after a database export, which is exactly why the chosen framework belongs in the registered methods and why our protocol development service fixes it before a single line is searched.
How to choose, and what comes next
Start from the question type. If you are comparing treatments, use PICO or PICOS. If you are mapping a field, reach for PCC. If you are synthesising experience or qualitative findings, SPIDER fits best. Whichever you choose, the framework becomes the skeleton of your research question and then of your eligibility criteria. There is no prize for using the most complex template; the best framework is the one that captures your question with the fewest forced elements.