GUIDE 1 OF 5

The foundation of rigor: systematic review protocols and PICO

A protocol is the contract you write before seeing results. Lock PICO (or SPIDER), eligibility, and synthesis methods so screening cannot drift. Then fill the free template and start a project.

Open Protocol template Start free project

Without a pre-defined protocol, methods tend to follow the results. That is how inclusion criteria quietly widen, outcomes get dropped, and a review becomes irreproducible. A protocol is not paperwork for its own sake: it is the logic that every later include/exclude call has to answer to.

Defining the research question with PICO and SPIDER

PICO (Population, Intervention, Comparison, Outcome) is the default for questions of clinical effectiveness. Write each element so two reviewers would mark the same abstract the same way. “Adults with diabetes” is still vague; “Adults (≥18) with type 2 diabetes, with or without established heart failure” can be tested against an abstract.

SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, Research type) is the better frame when the review is qualitative or mixed-methods: lived experience, implementation, or a phenomenon rather than a head-to-head intervention. A methodologist picks SPIDER when forcing PICO would fake a comparator that the literature does not have.

Put the chosen framework into the free protocol template before you write a search string.

The necessity of protocol pre-registration

PROSPERO (and similar registries) is a public contract: you stated the question, outcomes, and analysis before you saw the studies. That does not freeze science; it makes later amendments visible. Selective reporting of convenient outcomes is harder to hide when the original list is timestamped.

Register before screening starts. A registration filed after full-text decisions is not prospective, even if the form still accepts the record.

Establishing robust inclusion and exclusion criteria

Criteria should be yes/no tests, not essays. If two reviewers would argue about a borderline abstract, the criterion is still too vague.

Study design (RCT only, or also observational?)
Date range and language, if you will actually apply them
Setting and minimum follow-up, if they change eligibility
What happens to conference abstracts, protocols, and secondary analyses

Traceability — framework versatility

EvidenceFlow’s Question and Protocol stages (workflow steps 1 and 2) store this logic on the project. A later exclude is not a vibes-based skip; it cites a field you already defined.

Example trail

Stage	Logic / action
Step 1: SPIDER	Design: phenomenological qualitative studies only.
Step 4: Screening	Decision: Exclude — study design is an RCT.
Result	The exclusion reason is logged as study design for the PRISMA diagram.

Start a free project, paste this PICO, then continue with Guide 2: title and abstract screening.

FAQ

Is a protocol required for PRISMA?

PRISMA 2020 asks whether a protocol exists and where it was registered. Journals treat an unregistered, post-hoc protocol as a limitation.

PICO or SPIDER?

Use PICO when the question is about an intervention's effect. Use SPIDER when the question is about experiences, phenomena, or mixed designs that do not map cleanly onto an intervention-comparator pair.

Do students need PROSPERO?

If the review is intended for publication and meets PROSPERO eligibility, register before screening. Course-only exercises can stay as a local protocol.

Can I change the protocol later?

Yes, with a dated amendment and a reason. Silent changes after looking at results are data dredging.

Core import, screening, extraction, meta-analysis, and PRISMA reporting are free. AI relevance scoring and extraction auto-fill are the paid upgrade.

Use Protocol template

← All guides Guide 2: Screening efficiency →