Download the blank log and worked example

The blank CSV is ready to adapt; a separate CSV contains fictional records, and a TXT guide explains each field. The examples are not evidence or search results. No account is needed.

Use a blank record for your own work. Keep examples separate from your evidence and follow the instructions below.

A colorful fan of translucent study cards on a metal ring, with an evidence graph, highlighted passage, checklist and chart.
A visual reminder to keep research sources, notes and claims connected.
On this page

Your claim-to-source record

Use one record for one claim. Keep the original source open alongside it.

Notes stay in this tab and are not saved automatically. Download or print them before closing this page.

Copy the sentence you need to check.

Author, title, date, venue, DOI or URL.

Which publisher or library record did you check, and when?

Page, section or paragraph.

Include the words that limit the finding.

Who or what was studied? What was actually measured?

Describe any change in certainty, scope or meaning.

Write wording that the evidence can support.

Start with the protocol and version you are applying

A screening log records how a review team applies its eligibility rules to candidate records. It does not create those rules. Before screening, identify the approved protocol version, the question, population or context, eligible evidence types, date or language limits and any planned handling of borderline cases. Give each criterion a concise ID so a decision can point to the rule actually used.

If the protocol changes, preserve the earlier version and document what changed, when and why under your review procedure. Do not silently judge earlier records by a revised rule. A classroom literature search may need a simpler record; follow the assignment or supervisor instructions rather than presenting this worksheet as a systematic-review protocol.

Supporting reading: [2] [3]

Keep title and abstract decisions separate from full-text decisions

Use one row per record at each decision stage, or otherwise preserve a clear audit trail when the same record advances. At title and abstract screening, use the protocol’s permitted decisions. If the available summary does not answer an eligibility question, mark the record unclear and follow the planned rule for advancing it; do not infer full-text ineligibility from missing abstract detail.

At full text, record whether the report was obtained before recording an eligibility outcome. “Not retrieved” describes an access state. “Excluded” means an obtained report was assessed against a criterion and did not qualify. Keep the reason and the inspected section, page or metadata locator so another reviewer can understand the decision.

Supporting reading: [1] [2]

Preserve independent decisions and resolve disagreement transparently

Record each reviewer’s original decision before resolution if your review uses independent screening. Keep the resolution route specified in the protocol: for example, discussion, a designated adjudicator or another agreed process. Note who resolved the case and which criterion or evidence settled it. Do not replace two different decisions with a single unexplained label.

The number of reviewers and the conflict procedure depend on the review plan and its standards. This template does not make a particular reviewer arrangement mandatory. For a course project, a second person may be unavailable; record the method you actually used and disclose its limits accurately.

Supporting reading: [1] [2]

Worked example: unclear abstracts, access and a resolved decision

Fictional worked records for illustrating the fields only. These are not search results, publications, participants or review findings.

The downloadable example includes six constructed records: one duplicate removed before screening, one abstract advanced because population is unclear, one title/abstract exclusion, one full-text decision with a reviewer disagreement, one report still not retrieved, and one full-text exclusion with a protocol-linked reason. The labels and identifiers are illustrative only.

Try first: a database record has an abstract that does not state whether first-year results can be separated. The protocol permits mixed-year studies only when that group is separable. What should the title/abstract reviewer record, and what evidence is still needed?

Compare your choice with the worked CSV: mark the record unclear and advance it for full-text assessment under the planned rule. Do not exclude it merely because the abstract omits the detail. At full text, record the exact methods locator and apply the criterion. A second reviewer’s different decision remains visible until the documented resolution is complete.

Count records, reports and studies as different things

A database result is a record; a document assessed at full text is a report; one study can have multiple reports. Deduplication, screening exclusions, reports that could not be retrieved, full-text exclusions and included studies therefore belong to separate count fields. Do not add all row totals together as if each row represented a unique study.

PRISMA 2020 reporting asks authors to describe the selection flow, including records screened and excluded, reports sought and not retrieved, full-text exclusions with reasons, and studies and reports included. The official site provides different flow-diagram templates for different review and source configurations. Use the appropriate official template and your chosen reporting guideline; this worksheet is not a substitute and does not guarantee compliance.

When a retrieved report appears eligible but is excluded, retain its citation and explain the reason in the review report as applicable. Keep one primary reason per report if that is how the review’s flow counts are defined, while preserving any additional notes in the internal log.

Supporting reading: [1] [4]

Adapt the CSV without losing the audit trail

Retain a stable record ID and link multiple reports to the study they describe. Keep criterion IDs consistent with the protocol. Add fields required by your team, institution, funder or assignment, and restrict access if records contain personal or confidential information. Avoid collecting unnecessary sensitive data.

The blank CSV is a starting structure, not a validated data-management system. Pilot it on a small set of records, check that reviewers interpret the decision labels the same way, and revise the instructions before wider use. Keep an untouched copy of the original template and document any local changes.

Questions about this topic

What is a systematic review screening log?

It is a record of how candidate records and full-text reports were assessed against a review protocol, including decisions, reasons, evidence locators and any reviewer resolution procedure.

Should I exclude an article when its abstract is unclear?

Follow the review protocol. If the abstract does not establish a criterion, the planned process may advance it for full-text assessment. Do not treat missing abstract detail as proof of ineligibility.

Is a report I could not access an excluded study?

Not by itself. Record retrieval status separately from eligibility. PRISMA reporting distinguishes reports not retrieved from retrieved reports excluded after assessment.

Does this worksheet make my review PRISMA compliant?

No. It is an editable starting log, not a compliance check or protocol. Consult the current PRISMA guidance and the requirements for your review type and institution.

Sources and editorial notes

The supporting reading below informs our editorial guidance. Exercises and scenarios illustrate a method; they do not report a completed product test or measured learning outcome. Read our editorial policy or send a correction with a supporting source to info@ai-essay.com.

  1. PRISMA 2020: Official flow diagram templates
  2. University of Lincoln Library: Screening and selection
  3. University of Lincoln Library: The protocol
  4. PRISMA 2020 expanded checklist, items 16a–16b