01 · The Question
What if the Problem Is Not That the Search Is Old, but That It Was Weak?
Updating an existing literature review can reveal uncomfortable things about the original search. Perhaps an important synonym was missing. The Boolean logic was too restrictive. A key database was omitted. An inappropriate date, language, or publication-type limit was applied. Or the search was documented so poorly that you cannot confidently reconstruct what was done.
The easiest response is to leave the original strategy alone and use it again for consistency. But consistency with a flawed method simply reproduces the flaw. At the other extreme, starting the entire review from zero may discard substantial work that remains valid.
The more useful question is how consequential the weakness was, which part of the evidence base it could have affected, and how much of the search needs to be repaired.
02 · The Short Answer
Correct the Weakness and Investigate Its Historical Consequences
In Brief
If the original search strategy was weak, do not rerun it unchanged merely to preserve consistency. Identify the specific weakness, redesign the affected search components, test the revised strategy, and determine whether the problem could have caused relevant older studies to be missed.
If the weakness affected historical retrieval, applying the improved strategy only to newly published literature will not repair the original evidence base. Some or all of the earlier period may need to be searched again.
03 · What You Need to Know
Diagnose the Search Before Deciding How Much to Redo
Do not label a search "weak" without identifying the problem
A search can be imperfect without being unusable. Before redesigning it, specify what is wrong and how that weakness could affect retrieval.
PRESS provides a useful framework for examining electronic search strategies. Its major areas include whether the strategy translates the research question appropriately, uses Boolean and proximity operators correctly, selects suitable subject headings and text words, avoids spelling or syntax problems, and uses limits and filters appropriately. Structured peer review can identify errors and improve term selection.
Possible weakness
Potential consequence
Likely repair
Important synonym omitted
Relevant studies using different terminology may be missed.
Add appropriate terminology and assess the historical period.
Search concept too narrow
Eligible variants of the concept may be excluded.
Broaden the conceptual representation where justified.
Incorrect Boolean logic
Relevant records may be unintentionally excluded or retrieval may become distorted.
Correct the logic and retest the strategy.
Inappropriate limit or filter
Eligible records may never reach screening.
Remove or correct the restriction and search the affected period.
Important database omitted
Literature distinctive to that source may be underrepresented.
Add the source and determine how far back it should be searched.
Poor database translation
A strategy that worked in one source may retrieve inadequately in another.
Adapt vocabulary, fields, operators, and syntax correctly.
Incomplete documentation
The original search may not be reproducible.
Reconstruct what can be verified and clearly report remaining uncertainty.
Distinguish poor precision from poor coverage
A search that retrieves thousands of irrelevant records can certainly be inefficient, but that is not the same problem as a search that fails to retrieve eligible studies. Search quality involves trade-offs between recall and precision.
Low precision
The search retrieves many irrelevant records along with relevant ones, increasing screening burden.
Inadequate recall
The search fails to retrieve relevant records that should have had a reasonable opportunity to enter the review.
For an update, inadequate recall is particularly consequential because it can affect the composition and conclusions of the evidence base. An inefficient but adequately sensitive original strategy may need optimization rather than wholesale reconstruction.
Use the included studies as diagnostic evidence
Your existing review provides a useful set of records for testing the search. Check whether the original strategy retrieves known included studies that should be discoverable in the database being tested.
Garner and colleagues recommend using information about the yield of the previous review to refine update searches. Their supplementary guidance notes that original strategies can be optimized and that investigators should confirm that retained search approaches identify studies included in the original review.
Failure to retrieve a known included paper does not automatically prove that the entire strategy is invalid. The study may have been found through another source, or indexing may have changed. Repeated failures, however, can expose conceptual or technical problems worth investigating.
Look for errors that systematically exclude evidence
Some weaknesses are more consequential than others. A redundant synonym may make almost no difference. A misplaced Boolean operator can exclude an entire branch of relevant literature. An unjustified language restriction can systematically remove eligible studies. An incorrect study-design filter can do the same.
Ask what the error could have prevented from entering the candidate record set. This gives you a more defensible basis for deciding how much historical searching is necessary.
Do not repair only the future if the weakness affected the past
This is the central problem in updating a weak search. Suppose the previous review searched through December 2022. In 2026, you discover that a major synonym was absent. Adding that synonym to a 2023-2026 update improves future retrieval, but it does nothing about eligible studies from 2010-2022 that may also have used it.
If the weakness could have affected earlier records, search the affected historical period with the corrected approach. Depending on the nature of the error, that could mean rerunning one concept, one database, several sources, or a substantially redesigned search from the relevant starting point.
Watch Out
An improved update search can create the appearance of methodological progress while leaving the original evidence gap untouched. Always ask whether the correction needs to operate backward as well as forward.
Not every flaw requires rerunning everything
A targeted repair may be sufficient when the weakness is localized. If one database strategy contained a syntax error while all other sources were searched appropriately, you may need to repair that database rather than reconstruct the entire search.
Similarly, if a newly recognized term emerged only after the original search, there may be no historical gap to repair. This is why diagnosing the nature and timing of the weakness should come before deciding the search period.
Changed terminology may look like an original error when it is not
Be careful not to judge an old strategy using vocabulary that did not exist when it was developed. A search can have been reasonable at the time and still require revision now because the field has changed.
If the problem is primarily that terminology has changed since the original search , investigate when the newer terminology emerged before concluding that the earlier strategy was deficient.
An omitted database requires a coverage assessment
The fact that a database was not searched is not automatically evidence of a defective review. Appropriate database selection depends on the topic and eligible evidence.
If a relevant source was omitted, determine whether adding the database to the updated search could identify evidence absent from the original review and whether its historical holdings should also be searched.
Consider peer review before running the repaired strategy
If the weakness is substantial enough to require redesign, having the revised strategy checked by an information specialist can be valuable, particularly for systematic reviews. PRESS was developed specifically to support structured peer review of electronic search strategies and covers conceptual translation, operators, subject headings, text words, syntax, and limits.
This is especially useful before investing time in a large retrospective rerun. Discovering another search error after screening several thousand records is the methodological equivalent of finding a typo in the title after the journal sends the proofs.
Document the correction instead of rewriting history
Do not replace the original strategy in the review record and imply that the improved version was used previously. PRISMA-S was developed to support clear and reproducible reporting of search methods and information sources.
Report what the original search did, identify the weakness, describe the corrected strategy, state which dates or databases were searched again, and explain how newly identified historical studies were handled. Transparency is more credible than pretending the first version was methodologically pristine.
04 · A Practical Example
Repairing a Search That Used an Overly Restrictive Concept
Hypothetical Example
An updated review discovers that its original search was too narrow
Suppose a review examined online collaborative learning in higher education. The original strategy searched for "online collaborative learning" and a small number of close variants but omitted several common ways the eligible phenomenon was described. The final original search covered literature through December 2022.
Diagnose the problem
Testing shows that several known relevant records are not retrieved because the collaboration concept is represented too narrowly.
Redesign the affected concept
The researchers identify appropriate text words and controlled vocabulary while retaining useful original terminology.
Test the revision
The revised strategy is checked against known relevant records and reviewed for Boolean logic, syntax, subject headings, text words, and limits.
Determine the historical consequence
The missing terminology was already used before 2022, meaning the weakness could have affected the original evidence base.
Search backward
The corrected strategy is applied to the affected historical period as well as to literature published or indexed after the original search.
Process newly found older studies
Any newly retrieved historical records pass through the same appropriate eligibility and review procedures rather than being included automatically.
Report the repair
The update explains the limitation of the original strategy, what was changed, which period was re-searched, and whether the additional evidence affected the synthesis.
The update does not need to pretend the earlier search was adequate. Nor does it automatically require throwing away the entire previous review. The repair is proportional to the methodological problem.
06 · What This Means for You
Repair the Evidence Boundary, Not Just the Search String
A simple decision framework
If the weakness mainly reduced precision but relevant evidence was still adequately retrieved
Optimize the update search where appropriate without automatically repeating historical screening.
If one localized error could have excluded a specific group of older records
Correct that component and search the affected database and period.
If important terminology, concepts, or databases were missing throughout the original review
Redesign the affected search and consider substantial retrospective searching.
If the original search cannot be reconstructed reliably
Document the uncertainty and consider whether a new reproducible search across the required evidence period is necessary.
If correcting the search substantially changes the question, criteria, and evidence base
The objective is not methodological perfection after the fact. It is to identify whether the weakness could have distorted the evidence base and repair that risk as transparently and proportionately as possible.
This is also why rerunning the exact original search is not always the right approach. Reproducibility means documenting what was done. It does not require knowingly reproducing an avoidable error.
07 · A Quick Checklist
Before Repairing a Weak Original Search
Before rerunning the search, check:
Identify the specific weakness rather than describing the entire search vaguely as poor.
Check the translation of the review question into searchable concepts.
Review Boolean and proximity operators, subject headings, text words, syntax, limits, and filters.
Test whether known relevant and previously included studies can be retrieved where appropriate.
Determine whether the weakness affected precision, coverage, or both.
Identify the dates and databases across which eligible studies could have been missed.
Consider peer review of a substantially revised strategy before running a large retrospective search.
Search the affected historical period when the original weakness could have caused older evidence to be missed.
Preserve the original strategy and report the corrected version separately.
Assess whether newly recovered evidence changes the synthesis or previous conclusions.
09 · The Bottom Line
Do Not Preserve a Search Error for the Sake of Consistency
The Bottom Line
If the original search strategy was weak, diagnose and correct the specific weakness, then determine whether it could have caused eligible evidence to be missed during the original search period.
Repair only as much of the search as the problem requires, but do not limit the correction to new literature when the weakness also affected older evidence. Preserve the original search record, document the revised methods, and reassess the synthesis if the repaired search changes the evidence base.
11 · Cite this Guide
How to Cite This Guide
This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.
Recommended (Field Guide)
APA
MLA
Chicago
Copy Citation
