Screening Guide

Title & Abstract Screening

Best practices for the most time-consuming phase of your systematic review — and how to do it faster without sacrificing quality.

By Lumina Editorial Team Published Updated Methodology guide
Title and abstract screening interface with paper metadata and reviewer decisions
A screening interface should keep the citation, abstract, notes, and reviewer decision in one traceable workflow.

What Is Title & Abstract Screening?

Title and abstract screening is the first filter in a systematic review. You read each paper's title and abstract and decide whether it could be relevant based on your inclusion criteria. Papers that pass move to full-text review.

🎯 Key Principle: At this stage, be liberal. When eligibility cannot be determined from the available title and abstract, retain the record for full-text assessment rather than guessing.

How to Screen Effectively

1

Pilot Test Your Criteria

Screen the same calibration sample as a team. Discuss disagreements and refine your criteria before starting the full screening.

2

Read Title First, Then Abstract

Many papers can be excluded on title alone (e.g. clearly wrong population or topic). Only read the abstract when the title is ambiguous.

3

Use a Three-Decision System

Include (clearly relevant), Exclude (clearly irrelevant), Maybe (needs discussion or full text). The "Maybe" category reduces premature exclusions.

4

Screen in Batches

Use manageable sessions and monitor consistency. Session length should reflect abstract complexity, reviewer experience, and fatigue.

5

Document Everything

Retain decision history and counts. At full-text screening, record one primary exclusion reason for each excluded report for your PRISMA flow diagram.

Single vs. Dual Screening

Single Screening Dual Screening
How it works One reviewer screens all papers Two reviewers screen independently
Speed Fast Slower
Reliability Lower Higher
Best for Rapid reviews, scoping reviews Cochrane reviews, journal publications

Read our detailed Dual Screening guide → | Cochrane study-selection guidance

Using AI to Prioritize the Queue

Active learning can surface likely-relevant records earlier. That changes screening order; it does not by itself justify excluding unscreened records or replacing an independent reviewer.

Rank

Likely matches appear earlier

Check

Humans retain final decisions

Report

Document tools and stopping rules

Screen Smarter, Not Harder

Lumina's AI prioritizes the most relevant papers first, so you find what matters without reading thousands of irrelevant abstracts.