Title and Abstract Screening for Systematic Reviews

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-27

Quick Answer

• Title and abstract screening is the first structured filter that turns a large search result set into a smaller pool of records for full-text review.
• Build a screening form with explicit inclusion and exclusion criteria before you start, then test it on a sample of records to confirm consistency.
• Screening decisions are judgments, not proofs, so disagreements between reviewers are expected and must be resolved through discussion or a third reviewer.

Why Screening Quality Determines Review Outcomes

A systematic review depends on the quality of the records that survive the screening phase. If the screening step is too broad, reviewers waste time reading full texts that should have been excluded. If it is too narrow, relevant studies may be missed and the review conclusions may be incomplete or biased. The screening phase is where the search strategy and the final evidence base meet, and errors made here propagate through data extraction, synthesis, and interpretation.

The National Library of Medicine hosts research methods references that describe the role of structured screening in evidence synthesis. These references emphasize that screening is a replicable process, not an informal reading exercise. The goal is to apply the same decision rules to every record so that the review can be reproduced by another team.

Screening also matters for reporting transparency. The EQUATOR Network provides reporting guidelines that help authors describe how they selected studies. When you document the number of records identified, screened, excluded, and included, you give readers the information they need to assess the review's rigor.

Core Principles of Title and Abstract Screening

Screening Is a Two-Stage Filter

Most reviews use two screening stages. The first stage applies inclusion and exclusion criteria to titles and abstracts. The second stage applies the same criteria to full texts. The title and abstract stage is intentionally more permissive. If an abstract does not provide enough information to make a clear decision, the record moves to full-text review instead of being excluded prematurely.

This two-stage approach protects against premature exclusion. An abstract may omit details about study design, population, or outcomes that are described in the full text. Moving uncertain records forward is a deliberate strategy, not a failure of the screening process.

Inclusion and Exclusion Criteria Must Be Written Before Screening

Criteria should be written in the review protocol before screening begins. Each criterion should be specific enough that two independent reviewers can apply it consistently. Vague criteria such as "relevant studies" or "important outcomes" produce inconsistent decisions.

A useful screening form includes:

• A checklist of inclusion criteria that must all be met
• A checklist of exclusion criteria where any single match excludes the record
• A decision field for include, exclude, or uncertain
• A notes field for recording the reason for exclusion

The National Library of Medicine research methods references describe how explicit criteria support reproducible screening decisions. When criteria are written down and tested, reviewers can identify ambiguous cases before the full screening run begins.

Dual Review With Conflict Resolution

Independent dual review means that two reviewers screen each record without seeing the other reviewer's decision. This practice reduces the risk that one reviewer's interpretation of the criteria will dominate the process. After both reviewers have made their decisions, the results are compared.

Disagreements are resolved through discussion. If the two reviewers cannot agree, a third reviewer or the full review team makes the final decision. The Committee on Publication Ethics core practices describe the importance of transparent decision-making in research. Documenting how conflicts were resolved is part of that transparency.

Building a Screening Workflow

Step 1: Define the Research Question and Criteria

Start with a clearly defined research question. The question determines which populations, interventions, comparators, outcomes, and study designs are eligible. Write the inclusion and exclusion criteria directly from the question components.

For example, a review question about the effect of a specific feed additive on dairy cow milk production would specify:

• Population: lactating dairy cows
• Intervention: the specific feed additive at defined doses
• Comparator: standard diet without the additive
• Outcomes: milk yield, milk composition, or both
• Study design: controlled trials

Each of these components becomes a screening criterion.

Step 2: Build the Screening Form

Create a form that lists each criterion as a yes or no question. The form should also include the record identifier, the reviewer name, and the date. A notes field allows reviewers to record why a record was excluded, which supports later reporting.

The form can be a spreadsheet, a database, or a dedicated systematic review software tool. The format matters less than the consistency of the questions and the completeness of the records.

Step 3: Run a Calibration Exercise

Before screening the full set of records, select a random sample of 20 to 30 records. Have all reviewers screen the same sample independently. Compare the decisions and discuss every disagreement.

The calibration exercise serves two purposes. First, it identifies criteria that are ambiguous or difficult to apply. Second, it aligns the reviewers on how to interpret borderline cases. The EQUATOR Network reporting guidelines emphasize that screening methods should be described clearly enough for readers to understand what was done. A calibration exercise is part of that method description.

Step 4: Screen the Full Set Independently

Each reviewer screens every record independently. Do not split the records so that each record is seen by only one reviewer. Independent dual review is the standard because it provides a check on individual judgment.

During the full screening run, reviewers should record the exclusion reason for every excluded record. This record keeping supports the flow diagram that most reporting guidelines require.

Step 5: Resolve Conflicts

After both reviewers complete the full set, compare decisions. For each disagreement, the reviewers discuss the record and reach a consensus. If consensus is not possible, a third reviewer decides.

The conflict resolution process should be documented. The number of disagreements and how they were resolved is useful information for readers who want to assess the reliability of the screening process.

Step 6: Move to Full-Text Review

Records that pass the title and abstract stage move to full-text review. At this stage, reviewers read the complete article and apply the same criteria again. Some records will be excluded because the full text reveals that the study does not meet the criteria.

The National Library of Medicine research methods references describe the full-text stage as a separate and necessary step. Full-text review is not a formality. It is where the final inclusion decisions are made.

At a Glance

Options and Tradeoffs in Screening Tools

Spreadsheet-Based Screening

Spreadsheets are the simplest screening tool. Each row is a record, and each column is a criterion or decision field. Spreadsheets are free, widely available, and easy to customize.

The main limitation is scale. With thousands of records, spreadsheets become difficult to manage. Sorting, filtering, and tracking reviewer decisions across multiple reviewers becomes error prone. Spreadsheets also make it harder to enforce independent review because reviewers can see each other's decisions if they share the file.

Dedicated Systematic Review Software

Dedicated software packages are built for the screening workflow. They support independent review, conflict detection, and reporting. These tools typically generate the flow diagram data automatically.

The tradeoff is cost and learning time. Some tools require a subscription, and all require training before the team can use them efficiently. For a small review with a few hundred records, the learning time may not be worth the investment.

Reference Manager Screening

Some reference management tools include screening features. These tools are useful when the review team already uses the reference manager for other purposes. The screening features are usually less sophisticated than dedicated software, but they may be sufficient for a small review.

The choice of tool should match the size of the review, the number of reviewers, and the team's existing skills. The National Library of Medicine research methods references describe screening as a process that can be implemented with different tools, provided the process itself is rigorous.

Records and Measurements to Track

Screening Log

Maintain a screening log that records every decision. The log should include:

• Record identifier
• Reviewer name
• Screening date
• Decision for each criterion
• Final decision
• Exclusion reason

The screening log is the primary record of what happened during screening. It supports the flow diagram and provides the data needed to calculate inter-reviewer agreement.

Inter-Reviewer Agreement

Inter-reviewer agreement measures how often two reviewers make the same decision on the same record. Agreement can be calculated as a simple percentage, but this measure does not account for chance agreement. Cohen's kappa is a common statistic that adjusts for chance.

A low agreement score indicates that the criteria are ambiguous or that the reviewers are interpreting them differently. When agreement is low, the team should revisit the criteria and run another calibration exercise before continuing.

Flow Diagram Data

Most reporting guidelines require a flow diagram that shows the number of records at each stage. The flow diagram includes:

• Number of records identified through database searching
• Number of records after duplicates are removed
• Number of records screened
• Number of records excluded at title and abstract stage
• Number of full texts assessed for eligibility
• Number of full texts excluded with reasons
• Number of studies included in the review

The EQUATOR Network provides access to reporting guidelines that specify the flow diagram requirements for different review types. Keeping accurate counts at each stage makes the flow diagram straightforward to produce.

Common Failure Patterns in Screening

Screening Too Quickly

Reviewers who rush through records tend to make more errors. A record that takes only a few seconds to screen may not have been read carefully. The consequence is that relevant records are excluded or irrelevant records are moved forward.

A practical safeguard is to track screening time per record. If a reviewer is consistently screening records in under 30 seconds, the team should discuss whether the criteria are being applied carefully enough.

Excluding on Insufficient Information

Abstracts are often incomplete. A study may be relevant but the abstract may not mention the specific outcome or population of interest. Excluding such records at the title and abstract stage is a common error.

The rule should be that a record is excluded only when the abstract provides enough information to confirm that it does not meet the criteria. If the abstract is silent on a criterion, the record should move to full-text review.

Letting One Reviewer Dominate

When one reviewer is more experienced or more assertive, the other reviewer may defer to their judgment. This dynamic undermines the purpose of independent review. Each reviewer must record their own decision before any discussion occurs.

The screening process should be structured so that reviewers cannot see each other's decisions until both have completed the full set. This structure preserves independence.

Changing Criteria Mid-Screening

If the team discovers that a criterion is ambiguous, the temptation is to revise it and continue. Changing criteria mid-screening creates inconsistency because records screened before the change were judged under different rules.

If a criterion must be changed, the records already screened should be re-screened under the new criterion. This re-screening is time consuming, which is why the calibration exercise is important.

Poor Documentation of Exclusion Reasons

Exclusion reasons are needed for the flow diagram and for readers who want to understand why specific records were excluded. Reviewers who record vague reasons such as "not relevant" without specifying which criterion was not met create problems at the reporting stage.

The screening form should include a dropdown or checklist of exclusion reasons that map to the exclusion criteria. This structure forces reviewers to be specific.

Welfare and Safety Context in Life Science Screening

Animal Studies Require Careful Eligibility Assessment

Reviews that include animal studies must apply criteria that address animal welfare and study quality. The screening criteria should specify which animal species, models, and housing conditions are eligible. The National Institutes of Health provides guidance on the conduct of research involving animals, and reviewers should be aware of the standards that apply to the studies they are screening.

Screening criteria for animal studies should also consider whether the studies report welfare-relevant outcomes. A review about the effects of a housing system on pig health might include outcomes such as injury rates, disease incidence, and behavioral indicators. Studies that do not report these outcomes may still be relevant if they provide other data, but the criteria should be explicit about what is required.

Human Studies Require Ethics and Consent Verification

For reviews that include human studies, the screening criteria should address whether the studies report ethical approval and informed consent. Studies that do not report these elements may still be included if the review question does not require them, but the criteria should be explicit.

The Committee on Publication Ethics core practices describe the responsibilities of researchers in conducting and reporting research. Reviewers should be alert to studies that raise ethical concerns, and the review team should decide in advance how to handle such records.

Data Management During Screening

Screening generates a large amount of data, including decisions, notes, and exclusion reasons. The National Institutes of Health Data Management and Sharing Policy describes expectations for managing and sharing research data. While this policy applies to NIH-funded research, the principles of organized data management apply to any review.

The screening data should be stored in a format that is accessible to the review team and preserved for the duration of the project. The data should be backed up regularly to prevent loss.

Professional Escalation Criteria

When to Involve a Third Reviewer

A third reviewer should be involved when two reviewers cannot reach consensus on a record. The third reviewer should be someone who understands the review question and criteria but has not been involved in the initial screening of that record.

The third reviewer's decision is final. The outcome should be recorded in the screening log along with a note that the decision required third-reviewer input.

When to Revise the Protocol

If the calibration exercise reveals that a criterion is consistently misunderstood, the protocol should be revised before the full screening run begins. This revision is appropriate because it happens before any records have been formally screened.

If a criterion proves problematic during the full screening run, the team should stop and assess the situation. Continuing with a flawed criterion produces inconsistent data. The team should decide whether to revise the criterion and re-screen the affected records or to continue with the existing criterion and document the limitation.

When to Seek Methodological Advice

If the review team is uncertain about how to handle a specific screening situation, methodological advice may be needed. This advice might come from a librarian with systematic review expertise, a statistician, or a colleague with experience in evidence synthesis.

The National Library of Medicine research methods references provide background on screening methods, but they cannot address every specific situation. A methodological consultant can help the team apply the general principles to their specific review.

Limitations of Title and Abstract Screening

Abstracts Are Imperfect Representations

Abstracts do not always reflect the full content of a study. Some abstracts are poorly written, some omit key details, and some are misleading. The screening process cannot fully compensate for these limitations.

The two-stage approach mitigates this problem by moving uncertain records to full-text review. However, some relevant studies will still be excluded if their abstracts are so incomplete that the reviewers cannot recognize their relevance.

Screening Cannot Replace Full-Text Assessment

Title and abstract screening is a filter, not a final judgment. The full-text review is where the final inclusion decisions are made. Reviewers who treat the title and abstract stage as the final decision risk including records that fail criteria on full-text review or excluding records that would have passed.

The flow diagram should show the number of records excluded at each stage so that readers can see how the screening process worked.

Reviewer Judgment Is Inherent

Even with explicit criteria, screening involves judgment. Two reviewers may interpret the same abstract differently. The calibration exercise and conflict resolution process are designed to manage this variability, but they cannot eliminate it.

The review report should acknowledge this limitation and describe the steps taken to minimize its impact.

Frequently Asked Questions

What is the difference between title and abstract screening and full-text review?

Title and abstract screening is the first filter that removes obviously irrelevant records based on the information available in the title and abstract. Full-text review is the second stage where the complete article is read and the same criteria are applied with full information. Records that pass the first stage move to the second stage, and some records are excluded at the second stage when the full text reveals that they do not meet the criteria.

How many reviewers are needed for screening?

The standard is two independent reviewers. Each reviewer screens every record without seeing the other reviewer's decisions. Disagreements are resolved through discussion, and a third reviewer makes the final decision if consensus cannot be reached. Single-reviewer screening is faster but carries a higher risk of error and is generally not recommended for systematic reviews.

What should I do if the abstract does not provide enough information to make a decision?

Move the record to full-text review. The rule is that a record is excluded only when the abstract provides enough information to confirm that it does not meet the criteria. If the abstract is silent on a criterion, the record should be assessed at the full-text stage.

How do I calculate inter-reviewer agreement?

Inter-reviewer agreement is calculated by comparing the decisions of two reviewers on the same set of records. The simplest measure is the percentage of records where both reviewers made the same decision. Cohen's kappa is a more robust measure because it adjusts for agreement that would occur by chance. A low kappa value indicates that the criteria are ambiguous or that the reviewers are interpreting them differently.

What is a calibration exercise and why is it important?

A calibration exercise is a practice run where all reviewers screen the same sample of records independently and then discuss their decisions. The exercise identifies criteria that are ambiguous or difficult to apply and aligns the reviewers on how to interpret borderline cases. Skipping the calibration exercise increases the risk of inconsistent decisions during the full screening run.

Can I change my inclusion criteria after screening has started?

Changing criteria mid-screening creates inconsistency because records screened before the change were judged under different rules. If a criterion must be changed, the records already screened should be re-screened under the new criterion. This re-screening is time consuming, which is why the criteria should be tested thoroughly in the calibration exercise before the full screening run begins.

What records should I keep from the screening process?

Keep the screening log with every decision, the exclusion reasons for every excluded record, and the counts needed for the flow diagram. The screening log should include the record identifier, reviewer name, screening date, decision for each criterion, final decision, and exclusion reason. These records support the reporting of the review and allow the screening process to be audited.

Where can I find reporting guidelines for my review type?

The EQUATOR Network provides a searchable database of reporting guidelines for different study types and review types. These guidelines specify what information should be reported about the screening process, including the flow diagram and the number of records at each stage.

Using the Evidence

Related Articles

• How to Write a Scientific Paper Abstract: A Practical Guide
• Meta-Analysis vs Systematic Review: What's the Difference?
• How to Write a Science Fair Abstract: A Practical Guide for Students
• Scientific Literature Review: A Reproducible Search and Synthesis Workflow
• Meta-Analysis vs Systematic Review: Choosing the Right Synthesis Method

References and Further Reading

• Research Methods Resources. National Library of Medicine.
• EQUATOR Network. EQUATOR Network.
• Core Practices. Committee on Publication Ethics.
• NIH Grants and Funding. National Institutes of Health.
• ORCID for Researchers. ORCID.
• Data Management and Sharing Policy. National Institutes of Health.
• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.