How to Screen Articles for a Systematic Review: Title/Abstract to Full-Text

Screening is where systematic reviews win or fail. A weak search can miss studies, but a weak screening process can exclude the right studies by accident. This guide gives you a fast, reliable screening workflow from title/abstract to full-text, with PRISMA-ready documentation.

Helpful related reads: How to Build a PubMed Search Strategy, PRISMA 2020 Step-by-Step, How to Critically Appraise a Study.

What screening is (and what it is not)

Screening is a structured decision: include, exclude, or maybe. It is not a judgement of whether the paper is good. Quality appraisal happens later. Screening answers one question: Does this study meet the eligibility criteria?

Step 1: Lock your eligibility criteria before screening

Eligibility criteria should be written in measurable terms. If your criteria are vague, reviewers will disagree and your PRISMA flow will become messy. Use a PICO-based checklist and write:

• Population: age range, diagnosis definition, setting
• Intervention/Exposure: technique, drug, device, definition
• Comparator: standard care, other technique, or none
• Outcomes: what outcomes are required (if any)
• Study designs: RCTs only, or RCTs plus observational, etc.
• Timeframe and language limits: only if justified

If you are still refining the question, see: How to Write a Research Question.

Step 2: Use a 3-decision rule for title/abstract screening

At title/abstract stage, do not overthink. Use only three decisions:

• Include: clearly meets criteria or likely meets criteria
• Exclude: clearly fails criteria
• Maybe/Unclear: not enough information, move to full-text

Rule: If you are uncertain, do not exclude. Move it forward. False exclusion is more damaging than false inclusion at this stage.

Step 3: Dual screening and conflict resolution

Best practice is two independent reviewers screening titles/abstracts. Conflicts should be resolved by discussion or a third reviewer. This reduces selection bias and makes the review defendable in peer review.

Outbound reference for methodology: Cochrane Handbook.

Step 4: Define exclusion reasons only at full-text stage

Do not waste time assigning detailed exclusion reasons at title/abstract stage. PRISMA expects reasons for exclusions mainly at full-text stage. Create a short, standardized list of full-text exclusion reasons such as:

• Wrong population
• Wrong intervention/exposure
• Wrong comparator
• Wrong outcomes or follow-up window
• Wrong study design
• Duplicate or overlapping cohort
• Not original research (editorial, letter)

Keep reasons mutually exclusive and easy to audit.

Step 5: Full-text screening checklist (fast and accurate)

When reading the full text, do not read everything in detail. Use a structured scan:

1. Methods first: population, eligibility, and design
2. Intervention/exposure definition: match your protocol
3. Outcome and timeframe: confirm endpoints and follow-up
4. Duplicates: check cohort overlap, centers, dates

Only after it passes these checks should you invest time in deeper reading.

Step 6: PRISMA-ready documentation

PRISMA 2020 expects transparent reporting of the selection process, including a flow diagram. Record:

• Records identified per database
• Duplicates removed
• Records screened
• Full texts assessed
• Full texts excluded with reasons
• Studies included in qualitative and quantitative synthesis

Outbound reference: PRISMA Statement.

Step 7: Common screening pitfalls (and how to avoid them)

• Pitfall: excluding because outcomes are not mentioned in the abstract. Fix: move to full-text if population and intervention match.
• Pitfall: inconsistent application of criteria. Fix: do a calibration round on 50 records before full screening.
• Pitfall: duplicate cohorts included twice. Fix: track centers and recruitment years and choose the most complete dataset.
• Pitfall: overusing filters in searches then blaming screening. Fix: fix the search strategy first.

Internal workflow tip

If your platform supports it, keep screening decisions, reasons, and counts in one place so PRISMA numbers are automatic. Manage your projects here: Systematic Reviews Workspace and browse more methods posts on the SciTrack Blog.

Conclusion

Fast screening is structured screening. Lock eligibility criteria, use include/exclude/maybe at title/abstract stage, apply dual screening with conflict resolution, document full-text exclusion reasons, and keep PRISMA counts consistent. When you do this well, your review becomes faster, more accurate, and easier to publish.