Double Screening in Systematic Reviews
What double screening means, why it is also called dual screening, how two reviewers work independently, and how disagreements are resolved.
What Does Double Screening Mean?
Double screening, also called dual screening or independent screening, means two reviewers assess the same set of records independently, without seeing each other's decisions. Their decisions are compared afterwards, and disagreements are resolved through discussion or a third reviewer.
Double screening vs dual screening
There is no methodological difference between the terms. Both normally describe two people applying the same eligibility criteria independently. “Double screening” emphasizes that each record is assessed twice; “dual screening” emphasizes the two-reviewer workflow.
Reviewer A
Screens all papers independently
Reviewer B
Screens all papers independently
Conflict Resolution
Disagreements are discussed or adjudicated
Dual Screening Process Step by Step
1. Pilot the criteria
Both reviewers screen a small sample first so unclear inclusion and exclusion criteria can be fixed before the full review.
2. Screen independently
Reviewers make Include, Exclude, or Maybe decisions without seeing each other's labels.
3. Compare decisions
Agreements move forward automatically. Disagreements are marked as conflicts for review.
4. Resolve conflicts
Conflicts are resolved by discussion, a senior reviewer, or a predefined adjudication rule.
Why Is Dual Screening Important?
Reduces Error
Independent duplicate screening reduces the chance that one reviewer's fatigue, misunderstanding, or subjective interpretation determines study selection. The size of that benefit varies by review and workflow.
Minimizes Bias
Independent screening prevents one reviewer's biases (confirmation bias, familiarity bias) from dominating the selection process.
Supported by Methodology Guidance
The Cochrane Handbook describes duplicate independent title-and-abstract screening as desirable and requires at least two people for final full-text eligibility decisions in Cochrane intervention reviews.
Measurable Quality
Inter-rater reliability (Cohen's kappa) proves that your screening process was consistent and reproducible.
Measuring Agreement: Cohen's Kappa
Cohen's kappa (κ) measures the level of agreement between two reviewers, adjusting for chance. Report this in your methodology section.
| Kappa (κ) | Agreement Level | Interpretation |
|---|---|---|
| < 0.20 | Poor | Criteria need refinement — run another pilot |
| 0.21–0.40 | Fair | Some disagreement — discuss and clarify criteria |
| 0.41–0.60 | Moderate | Acceptable for most reviews |
| 0.61–0.80 | Substantial | Good agreement — proceed with confidence |
| 0.81–1.00 | Almost Perfect | Excellent — criteria are very clear |
Single Screening vs Double Screening
Double screening reduces reliance on one person's judgement but requires more reviewer time. Whether a single-reviewer approach is acceptable depends on the review type, protocol, publisher, and applicable methodology guidance. Cochrane, for example, allows one person to perform the initial title-and-abstract screen but requires two independent people for final full-text eligibility decisions in intervention reviews.
Some rapid-review protocols use one screener with verification of exclusions; document and justify any shortcut.
AI can order the screening queue or support checking, but it should not be described as an independent human reviewer unless the protocol explicitly validates that approach.
If full double screening is not feasible, predefine which stage is checked by a second person and report the limitation transparently.
Dual Screening in Lumina
Lumina supports blinded dual screening out of the box. Two team members screen independently, and conflicts are flagged automatically for resolution.
- ✅ Blinded mode — reviewers can't see each other's decisions
- ✅ Automatic conflict detection
- ✅ Built-in conflict resolution interface
- ✅ AI still ranks papers for both reviewers