Is This Systematic Review Still Current?

Key points#

• Locate the last search date for every database, registry, website, or other source. The article’s publication date does not define evidence coverage.
• Ask whether the question and methods still fit current practice before adding newer studies to an older synthesis.
• New evidence matters when it changes an effect estimate, its precision, certainty, applicability, or understanding of benefit and harm.
• A living systematic review requires continuing surveillance and version control. The word “living” is not evidence that the latest search has been incorporated.
• A quick bridge search can identify candidate new studies, but it is not itself a systematic update.

Find the evidence boundary#

Every systematic review has a point after which evidence is absent. That boundary is set by the dates on which its information sources were last searched or consulted. PRISMA 2020 asks authors to identify each source and its last search date so that you can judge completeness and currency.

Several dates may appear in one paper:

• the protocol registration date;
• the final search date for each source;
• the manuscript submission, acceptance, and publication dates;
• the date of a correction or revised version;
• the date on which a living-review surveillance cycle found no eligible new evidence.

These dates answer different questions. Publication tells when the report became available. The search date tells how far the evidence collection reached. A correction may fix the text without adding studies. A surveillance note may confirm that a recent search found nothing eligible, even though the meta-analysis itself did not change. So read the methods and the supplementary search strategies rather than the cover page: if the abstract says “through March” but individual databases were searched on different days, the source-specific dates are the precise record.

Age is not a verdict#

There is no universal number of months after which every review becomes unusable. The Cochrane update framework instead asks whether the question remains important and well framed, whether the existing review used valid methods, and whether new studies, information, or methods could affect the findings.

A mature topic with many consistent, precise trials may change little after one additional small study. A synthesis based on two uncertain trials may move substantially when a larger study arrives. Some fields change because the intervention, comparator, diagnostic definition, or background care has changed, even if the pooled arithmetic remains intact.

Currency therefore has at least four dimensions:

1. Evidence currency: Have eligible results appeared beyond the search boundary?
2. Question currency: Do the population, intervention, comparator, and outcomes still represent the decision?
3. Method currency: Would current risk-of-bias or synthesis methods materially change the analysis?
4. Integrity currency: Have included studies been corrected, retracted, or supplemented with important new data?

An old date raises a question. It does not supply the answer.

Run a reader’s bridge check#

You can run a focused check without pretending to reproduce the review.

Reconstruct the original question#

Write the eligibility criteria in one sentence. Include population, intervention or factor, comparator, outcomes, study designs, and any setting or time restrictions. Without this frame, a search may retrieve newer papers that look relevant but would never have qualified.

Verify where and when the authors searched#

Check bibliographic databases, trial registries, regulatory sources, citation searching, and any specialized repositories appropriate to the topic. A recent date in one database does not compensate for an old or missing search elsewhere. Compare the reported strategy with the review’s claim of comprehensiveness.

Look for candidate evidence after the boundary#

Search the review title or key question terms together with “randomized,” a relevant study-design term, and a date filter beginning after the final search. Search trial registrations named as ongoing in the review. Use citation tools to find papers that cite the review. This is a bridge check, not a formal update: screening a few search results cannot establish that nothing was missed.

Check the status of included evidence#

Look for corrections, expressions of concern, retractions, longer follow-up, secondary outcome reports, or newly released registry results. A change in an existing study can matter as much as a newly recruited trial.

Estimate whether the conclusion is vulnerable#

Read the certainty assessment, confidence intervals, heterogeneity, risk-of-bias judgments, and contribution of each study. One new trial is more likely to matter when the original estimate is imprecise, one study dominates the weighting, important outcomes were missing, or the certainty was low. What you end up with should be modest: “candidate new evidence exists,” “the original question appears to have changed,” or “no obvious trigger was found in a limited check.” Only a planned, reproducible update can replace the review’s evidence set.

Decide what kind of change is needed#

An update is not always a matter of appending a few rows to a forest plot. The review team should first revisit the background, objectives, eligibility criteria, and analytical methods.

Minor changes may include adding eligible studies under the same protocol and repeating the planned analyses. More substantial work may be needed when outcome definitions have changed, new comparators have become standard, risk-of-bias methods have advanced, or the original synthesis method is no longer suitable. If the question or methods change enough, a new protocol and a new review may be clearer than presenting the work as a continuation.

New evidence can affect a synthesis in several ways:

• shift the estimated direction or size of an effect;
• narrow or widen uncertainty;
• change certainty without moving the central estimate;
• add information about harms or patient-important outcomes;
• reveal variation across settings or groups;
• make the original comparison less applicable to current care.

“The conclusion did not reverse” is therefore too narrow a definition of an important update. Increased certainty, newly recognized harm, or reduced applicability can alter decisions even when the pooled point estimate looks similar.

What living mode actually requires#

Cochrane defines a living systematic review as a review that is continually updated as relevant new evidence becomes available. The method is most defensible for a high-priority question, uncertain existing evidence, and a reasonable expectation that new studies will appear.

A credible living protocol should state:

• how often each source will be searched;
• how automated and human screening will be combined;
• when newly eligible evidence will be incorporated;
• how repeated analyses and statistical error will be handled;
• how readers will see what changed between versions;
• who is responsible for surveillance and quality control;
• what conditions will end, pause, or change the living process.

The landing page should show the latest search and version status. A journal label without a recent search record can give false reassurance. Conversely, a cycle that finds no new eligible study can still be a meaningful update if the date and result are reported transparently.

Living methods also create operational demands. Search results must be screened repeatedly, conflicts resolved, data extracted, bias appraised, analyses checked, and downstream summaries synchronized. Automation can help prioritize records, but accountable humans must maintain the eligibility decisions and conclusions.

Read currency as a documented claim#

Use a systematic review only for what its evidence boundary supports. Record the last search date, note any candidate later studies, and separate a limited currency check from a formal update. Cochrane’s current guidance expects review searches to be recent enough for publication and provides specific reporting conventions, but readers still need to inspect the actual dates and methods.

The most informative sentence is not “this review is recent.” It is: “These sources were searched through these dates, newer evidence was assessed by this process, and this version reflects the result.” The PRISMA diagram guide can help evaluate how the reported evidence set was assembled before currency is considered.