Keep Unavailable Full Texts Separate from Ineligible Studies

Preserve the distinction

A report may appear relevant at title and abstract screening but remain inaccessible. That is a retrieval problem. It does not show that the underlying study fails the eligibility criteria. Create an explicit awaiting-retrieval or awaiting-classification status rather than assigning a convenient scientific exclusion reason.

Cochrane’s selection guidance describes a staged process from records to reports and studies. Keep these units distinct when accounting for inaccessible material so the flow does not imply that every record received a full-text eligibility assessment.

Make retrieval reproducible

Record the full citation, identifiers, source database and attempted routes. Check publisher access, authorized institutional services, repositories, linked reports and appropriate author contact. Preserve dates and outcomes. Do not bypass access controls or assume that any file with a matching title is the correct final report.

If an alternative version is obtained, record its status and compare its contents with the citation. A conference abstract may identify a study but still lack information needed to determine eligibility or extract an outcome. Another article from the same study may resolve the question without replacing the missing report itself.

Reconcile a hypothetical flow

Imagine 80 reports are sought after initial screening. Seventy-four are obtained and six remain unavailable at the cutoff. Of the 74 obtained reports, 50 are excluded on documented eligibility grounds and 24 describe eligible studies.

The six unavailable reports do not belong among the 50 full-text exclusions. Report the retrieval gap separately. The 24 eligible reports may describe fewer than 24 studies, so link study families before stating the final study count. If two unavailable reports are later recovered, update their status and preserve the earlier flow version.

Explain the consequence

Describe whether missing reports could plausibly change the evidence base, using what is actually known from abstracts, registries or linked publications. Avoid declaring the impact negligible simply because the reports are difficult to obtain. Equally, do not infer unfavorable results from lack of access.

Set a transparent cutoff and maintain a follow-up list for later updates. If a report cannot be classified, state which eligibility information is missing. The final record should distinguish retrieval completed, eligibility assessed and outcome data available. Those are three different achievements, and a review becomes easier to trust when its flow diagram and extraction database preserve all three.

Source verification: 6 November 2026. Examples are hypothetical.