revisia Log in Sign up
Explore the guide

User guide · Stage 4 of 10

Criteria

Decide which studies count. revisia applies criteria in practice: the AI cannot include a study that fails a required criterion.

Open revisia
Video for Criteria · 2:03 · recorded in septiembre de 2026, so the interface may have changed slightly. Audio is in Spanish. See all ten.
Why criteria come after the search

You write criteria once you can see the actual set of titles and abstracts. This shows what can and cannot be checked and prevents criteria that sound good but cannot be applied.

Define the criteria

  1. Start from a template or create your own. Use a built-in set, choose “Suggest criteria (AI)” or add them manually. You always confirm every criterion before screening.
  2. Make them specific and testable. “Adults (≥18 years)” can be checked; “appropriate population” cannot. Also state what to do when the abstract does not mention the information, which is the most common case.
  3. Decide whether it is required. Failing a required abstract criterion excludes a study. Optional criteria provide information but do not force that decision.
  4. Check that at least one criterion is for abstracts. Screening cannot start without one.

Abstract or full text

Each criterion has a stage label. This is the most important decision on this screen.

  • “Abstract” is for facts that can be checked from the title and abstract. These criteria are applied during stage 5 screening.
  • “Full text” is for details that require the whole article, such as the randomisation method, exact dose or measurement instrument.
Rule of thumb

If the answer is not in the abstract, it is a full-text criterion. Forcing it into the first phase creates exclusions that you will later have to rescue one by one.

Troubleshooting

  • Screening asks for an abstract criterion. Return here and mark at least one criterion as “Abstract”.
  • I do not know where to start. Built-in sets contain common criteria for each review type. Remove what does not apply and add your own requirements.
  • One criterion excludes too much. It often fails to say what to do when the abstract is silent. Add “if not reported, mark uncertain” so doubtful records go to your queue instead of being excluded.
  • I want to change a criterion after screening. You can, but screened studies are not recalculated automatically. Run screening again so decisions reflect the new rule.
Still stuck?

Ask the assistant inside your project or email us at equipo@revisia.es.