How to Conduct Full-Text Screening

Full-text screening is the second selection stage: you read each potentially eligible paper in full, apply the inclusion criteria and record a single main reason for each exclusion. It decides which studies enter the review.

Steps

1. Retrieve full texts through library access, open-access sources and author contact; record those you cannot retrieve.
2. Apply the criteria in the same order using a short form with one question per criterion (population, intervention, comparator, outcome, design).
3. Screen independently in duplicate and compare.
4. Record one primary reason for each exclusion, using a fixed list of reasons.
5. Link multiple reports (for example the main paper, a protocol, and a follow-up) to the same study.
6. Resolve disagreements and finalise included studies.
7. Update the flow diagram counts.

Practical points

• The same trial may be reported in several papers; count the trial once and combine information.
• If data you need are missing, contact the authors and record what you asked and when.
• Keep a list of excluded studies with reasons for the supplement.

Illustrative example

Invented for teaching. A review includes randomised trials of vitamin D for tuberculosis. Of 60 full texts, 18 are excluded: 7 not randomised, 5 wrong population, 4 no relevant outcome, 2 duplicates of another report. Two reports describe the same trial, so the review has 41 reports of 38 trials.

When full texts are hard to get

Try your library, interlibrary loan, open-access repositories and the corresponding author. Record how many reports could not be retrieved and discuss the possible effect in the limitations. Do not treat an abstract alone as sufficient unless your protocol allows it, because decisions on abstracts may be wrong for borderline studies.

Common pitfalls

• Excluding for several reasons without ranking them.
• Counting reports as studies.
• Not retrieving some full texts and not reporting this.

Key takeaways

• Work in duplicate, record reasons and link reports.
• Keep the counts and the list of exclusions.

This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.