The search is the one stage of a systematic review whose errors cannot be recovered later. A study you never retrieve is a study you can never screen, appraise, extract, or synthesize, and no amount of methodological care downstream can compensate for it. This is why the search strategy is the single most consequential thing to get right, and why an inadequate search is the most common fatal flaw reviewers and editors find. Yet most researchers build a search, run it, and only discover it was flawed when a peer reviewer at a journal, or an examiner at a viva, points to the gap months later. There is an established solution to this, and it exists precisely because search errors are both common and preventable: peer review of the search strategy itself, before it is run, against the PRESS guideline. Our systematic review writing service builds searches to this standard, but every researcher should understand it, because it changes when and how you validate your search.
This guide sets out what PRESS is, the six elements a search peer reviewer assesses, why the review happens before the search runs, how PRESS differs from PRISMA-S, and what to do when a journal tells you your search is inadequate.
Quick Answer:
PRESS (Peer Review of Electronic Search Strategies) is a structured process in which an information specialist reviews your systematic review search strategy against six elements before you run the definitive search: translation of the research question, Boolean and proximity operators, subject headings, text word searching, spelling and syntax, line errors, and limits and filters. It matters because search errors are common and irrecoverable: one study found design problems in 73 percent of Cochrane review searches, and another found errors in 92.7 percent of a sample of published reviews. PRESS is a peer review of the search strategy, a process done before the search runs. It is distinct from PRISMA-S, which is a reporting checklist used at write-up. The Cochrane Handbook strongly recommends that all search strategies be peer reviewed before being run.
Why the Search Is the Stage You Cannot Afford to Get Wrong
Every other stage of a systematic review is, in principle, recoverable. If your screening is too strict, you can re-screen. If your data extraction misses a variable, you can go back to the papers. If your risk of bias judgments is questioned, you can reassess. The search is different. If your strategy fails to retrieve a relevant study, that study is simply absent from every subsequent stage, and nothing downstream will surface it. The error propagates silently through the entire review and biases the final synthesis, and you will not know it happened.
The empirical picture is sobering. Franco and colleagues, examining a random sample of Cochrane intervention reviews, found problems in the design of the search strategies in 73 percent of reviews, and more than half of those contained problems serious enough to limit both the sensitivity and the precision of the search (Research Synthesis Methods, 2018). Salvador-Oliván and colleagues, analyzing search strategies in 137 systematic reviews, found that the proportion containing some type of error was 92.7 percent, with errors affecting recall the most frequent (Journal of the Medical Library Association, 2019). These are not fringe reviews. Cochrane reviews are among the most methodologically scrutinized in the field, and the error rate is still high. The lesson is not that researchers are careless; it is that search construction is genuinely difficult, and errors are invisible to the person who made them, which is exactly the condition under which independent peer review is most valuable.
What PRESS Is, and Where It Came From
PRESS stands for Peer Review of Electronic Search Strategies. It is an evidence-based guideline for the peer review of the search strategies used in systematic reviews, developed by a team led by Jessie McGowan and Margaret Sampson and published as the 2015 Guideline Statement (McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. "PRESS Peer Review of Electronic Search Strategies: 2015 Guideline Statement." Journal of Clinical Epidemiology 2016;75:40-46). It grew out of an earlier 2009 evidence-based guideline by the same group and was updated through a systematic review of the evidence, a web-based survey of expert searchers, and a consensus process. A companion Explanation and Elaboration document, published by CADTH, gives detailed guidance on the submission and assessment form.
The central idea is simple and powerful: the search strategy is a technical artifact that can be checked against known failure modes, and a second qualified pair of eyes catches errors the original searcher cannot see. PRESS turns that principle into a structured, repeatable process built around six assessment elements.
The Six PRESS Elements a Reviewer Assesses
The 2015 update retained six of the seven elements from the original guideline; the seventh, concerning the skilled translation of the strategy to additional databases, was removed by consensus because it was judged best left to the searcher's discretion. The six that remain are the checklist a PRESS reviewer works through.
Translation of the research question. Does the search strategy correctly render the research question, and its PICO or equivalent concepts, into search terms and structure? The reviewer checks that each concept in the question is represented and that no concept has been dropped or misunderstood. A search that does not faithfully translate the question cannot answer it, however well-built the rest of the strategy is.
Boolean and proximity operators. Are AND, OR, and NOT used correctly, and would precision improve with proximity or phrase searching? The classic error here is using AND where OR is required, which silently narrows the search and drops relevant records. The reviewer checks the logic of how concepts are combined and whether proximity operators (adjacency, near, within) are used appropriately and with the right width.
Subject headings. Are the database-specific controlled vocabulary terms, MeSH in MEDLINE, Emtree in Embase, and their equivalents, used adequately and appropriately? The reviewer checks whether relevant headings have been included, whether explosion has been used where it should be, and whether any restriction to major headings or focused terms is justified, since over-restriction silently loses records.
Text word searching. Beyond the controlled vocabulary, has the searcher captured the free-text terms: the synonyms, spelling variants, acronyms, and abbreviations, and appropriate truncation? Subject headings alone never suffice, because indexing lags and is inconsistent, so a robust search combines headings with a thorough set of free-text terms. The reviewer checks for missing synonyms and for truncation that is either too aggressive or too timid.
Spelling, syntax, and line numbers. Are there misspellings, database syntax errors, or incorrect line-number references in combined statements? These are the mechanical errors that are trivially easy to make and surprisingly easy to miss: a misspelled term retrieves nothing, a syntax error breaks a line, and a wrong line reference in a combination step can silently corrupt the entire strategy. The reviewer checks the strategy line by line.
Limits and filters. Are the limits (date, language, publication type) and search filters justified, correctly constructed, and free of bias? An inappropriate limit, a language restriction that excludes eligible studies, or a poorly validated filter can introduce systematic bias into what the search retrieves. The reviewer checks that every limit is defensible and that any filter used is a validated one applied correctly.
Table 1: The Six PRESS 2015 Assessment Elements
Element | What the Reviewer Checks | Classic Error It Catches |
|---|---|---|
Translation of the question | Every PICO concept is represented in the strategy | A concept dropped or misunderstood |
Boolean and proximity operators | Correct AND/OR/NOT logic; proximity where useful | AND used where OR is required, collapsing the search |
Subject headings | MeSH/Emtree adequate, exploded, justified restrictions | Missing headings or unjustified major-heading limits |
Text word searching | Synonyms, variants, acronyms, truncation | Reliance on headings alone; missing free-text synonyms |
Spelling, syntax, line numbers | No misspellings, syntax errors, or wrong line references | A wrong line reference corrupting a combination step |
Limits and filters | Limits justified; validated filters applied correctly | A language limit that excludes eligible studies |
Why the Review Happens Before You Run the Search
The timing is the part researchers most often get wrong, and it is the part that matters most. PRESS is not a check you run on a completed search to report in your manuscript. It is a quality-improvement step performed before the definitive search is executed, so that errors are caught and corrected before they contaminate your results. The 2015 expert consensus placed the review at a specific point: after the primary database strategy, typically MEDLINE, has been built, but before it is translated to the other databases and run. Reviewing at this point means an error in the core strategy is fixed once, rather than being faithfully replicated across every database you then search.
This timing is now embedded in methodological standards. The Cochrane Handbook, in its chapter on searching for and selecting studies, strongly recommends that all search strategies be peer reviewed before they are run by a suitably qualified and experienced medical or healthcare librarian or information specialist. Cochrane's methodological expectations (MECIR) reinforce this, requiring that the search process be documented in enough detail to be reproduced and that searches be current at the time of publication. The point of running the peer review before the search, rather than after, is that a search reported transparently but built incorrectly is still a broken search. Getting the timing right is what makes the review preventive rather than merely descriptive, and it pairs naturally with pre-registering your plan, which our guide on writing a systematic review protocol and registering it on PROSPERO covers in full.
Who Should Peer Review Your Search
The reviewer should be an information specialist or medical or healthcare librarian experienced in systematic searching, someone who builds and critiques these strategies as part of their professional practice. This is a specialist competency, distinct from subject expertise: a content expert can tell you whether a concept is missing, but only an experienced searcher can tell you whether your Emtree mapping is right, whether your proximity width is optimal, or whether your MEDLINE strategy will translate cleanly to Embase syntax. The 2015 guideline is explicit that both the person requesting the review and the person performing it should be experienced, because PRESS is a peer process between people who both understand searching. If your institution has a health sciences librarian, that is your first port of call. Where in-house expertise is not available, this is one of the specific, modular points at which researchers bring in external support, and it is a service we provide as a discrete piece of work rather than only as part of a full review.
PRESS Versus PRISMA-S: The Distinction Competitors Blur
This is the distinction that separates a researcher who understands search methodology from one who does not, and many online guides get it wrong by treating the two as interchangeable. They are complementary, but they do different jobs at different times.
PRESS is a peer review of the search strategy. It is a process, performed before the search runs, whose purpose is to find and fix errors so the search is sound. Its output is a corrected strategy.
PRISMA-S is a reporting checklist. It is the PRISMA extension for reporting literature searches (Rethlefsen ML, Kirtley S, Waffenschmidt S, et al. "PRISMA-S: an extension to the PRISMA Statement for Reporting Literature Searches in Systematic Reviews." Systematic Reviews 2021;10:39), and it comprises 16 reporting items covering how to describe the search transparently and reproducibly at write-up: which databases and registries were searched, the full strategy for each, the dates, the limits, and how records were managed. Its purpose is transparency and reproducibility, not error detection.
The relationship is straightforward once you see it. PRESS makes the search correct; PRISMA-S makes the search reportable. You use PRESS before you run the search to fix it. You use PRISMA-S when you write up, to describe it. The two even reference each other: PRISMA-S includes an item asking authors to report whether the search was peer reviewed, which is to say, whether a PRESS-style review was done. Conflating them, using PRISMA-S as if it validated the search, or claiming a search is sound because it was reported to PRISMA-S, is a category error, because a strategy can be reported perfectly and still be built wrongly.
Table 2: PRESS vs PRISMA-S — Two Standards, Two Jobs
Dimension | PRESS (2015) | PRISMA-S (2021) |
|---|---|---|
What it is | Peer review of the search strategy | Reporting checklist for the search (16 items) |
Purpose | Find and fix errors | Transparency and reproducibility |
When it is used | Before the search is run | At write-up |
Output | A corrected strategy | A transparent, reproducible search report |
What it makes the search | Correct | Reportable |
The Errors PRESS Reliably Catches
It helps to see the concrete failure modes, because they are specific and recurring. A PRESS review reliably catches missing spelling and synonym variants, where a searcher has included one term for a concept but not its common alternatives. It catches Boolean logic errors, most damagingly, AND used where OR belongs, which silently collapses a search. It catches missing subject headings and the opposite error of over-relying on MeSH without the free-text terms needed to catch recent or poorly indexed records. It catches biased or unjustified limits and filters, such as a language restriction that quietly excludes eligible studies. It catches line-reference errors in combined statements, where a strategy built from numbered lines combines the wrong ones. And it catches truncation and proximity mistakes, where a wildcard is too broad, too narrow, or misplaced. The CADTH Explanation and Elaboration report, documenting the errors found in its own sample, listed missing subject headings, omitted free-text synonyms, and errors in syntax or Boolean operators as the most common. These are precisely the errors that are invisible to the person who made them and obvious to a second qualified reader, which is the entire rationale for peer review. Search quality feeds directly into whether a review survives peer review at all, a theme our guide on why systematic reviews get rejected develops across the whole manuscript.
Want your search strategy peer-reviewed before you run it? |
|---|
Send us your research question and your draft strategy. An information specialist trained in PRESS will assess all six elements, catch the Boolean, subject-heading, and translation errors before they contaminate your results, and return a corrected, reproducible strategy documented for PRISMA-S. Request a search strategy review and receive an itemized quote within 2 to 4 business hours, no obligation. |
How to Get a PRESS Review
The process is practical and worth setting up deliberately. Build your primary-database strategy first, usually in MEDLINE, to the point where you consider it complete: all concepts translated, subject headings and free-text terms combined, limits set. Assemble the materials the reviewer needs: the research question and its PICO concepts, the eligibility criteria, and the full strategy exactly as it will run, with line numbers. Then submit it to a qualified information specialist using the PRESS submission form, which structures the request around the six elements. The reviewer works through each element, marking each as requiring no revision, revision suggested, or revision required, and returns the annotated strategy. You correct the strategy, and only then translate it to the other databases and run the definitive search. Document that the peer review took place, because you will report it under PRISMA-S when you write up. Building the search correctly from the start also makes the appraisal and synthesis stages cleaner, which our guide on critically appraising studies in a systematic review picks up once the studies are in.
What to Do When a Reviewer Says Your Search Is Inadequate
This is a common and recoverable situation, and the response should be systematic rather than defensive. First, map the criticism onto the six PRESS elements: an "inadequate search" comment almost always points to one or more of them, whether a missing concept (translation), a Boolean error, missing subject headings or free-text terms, or an unjustified limit. Second, if you have not already had a formal PRESS review, obtain one now, so your revision is grounded in the guidelines rather than in guesswork. Third, re-run the corrected strategy and document every change you made, so the revision is transparent. Fourth, report the full strategy for every database, with dates and record counts, per PRISMA-S, since reviewers frequently raise search adequacy precisely because the reporting was too thin to judge. Fifth, if the scope of retrieval changed materially, note any re-screening. Framing your response as a documented, standards-based revision, PRESS to fix, PRISMA-S to report, is far more persuasive to an editor than an ad hoc patch, and it is often the difference between acceptance and a second round of major revisions. This modular, single-stage support, fixing and revalidating a search after a reviewer challenge, is one of the most common points at which researchers reach us, alongside our related scoping review and meta-analysis services.
Frequently Asked Questions
What is PRESS in a systematic review?
PRESS stands for Peer Review of Electronic Search Strategies. It is an evidence-based guideline, published as the 2015 Guideline Statement by McGowan and colleagues, for having an information specialist review a systematic review search strategy before it is run. The reviewer assesses the strategy against six elements to catch errors that would otherwise propagate through the entire review. It is the recognized standard for validating a search strategy prior to execution.
What are the six PRESS elements?
The six elements are: translation of the research question into search terms, correct use of Boolean and proximity operators, adequate use of subject headings such as MeSH and Emtree, thorough text word or free-text searching, freedom from spelling, syntax, and line-number errors, and justified limits and filters. A PRESS reviewer works through each element in turn, marking each as needing no revision, revision suggested, or revision required, and returns an annotated strategy for correction.
When should a search strategy be peer reviewed?
Before the definitive search is run. The consensus point is after the primary database strategy, usually MEDLINE, has been built, but before it is translated to other databases and executed. Reviewing at this point means a core error is fixed once rather than replicated across every database. The Cochrane Handbook strongly recommends that all search strategies be peer reviewed before being run by a qualified librarian or information specialist.
What is the difference between PRESS and PRISMA-S?
PRESS is a peer review of the search strategy, a process done before the search runs to find and fix errors; its output is a corrected strategy. PRISMA-S is a 16-item reporting checklist used at write-up to describe the search transparently and reproducibly. PRESS makes the search correct; PRISMA-S makes it reportable. They are complementary, and PRISMA-S even asks authors to report whether the search was peer reviewed. Treating PRISMA-S as if it validated the search is a category error.
Who should peer review my search strategy?
An information specialist or medical or healthcare librarian experienced in systematic searching. This is a specialist competency distinct from subject expertise: a content expert can spot a missing concept, but only an experienced searcher can judge subject-heading mapping, proximity width, and cross-database translation. The 2015 guideline expects both the requester and the reviewer to be experienced, because PRESS is a peer process between people who both understand searching.
How common are errors in systematic review searches?
Very common. One study found design problems in 73 percent of a sample of Cochrane review searches, with more than half of those serious enough to limit sensitivity and precision. Another found some type of error in 92.7 percent of search strategies across 137 reviews, with recall-affecting errors the most frequent. Because Cochrane reviews are among the most scrutinized in the field, these figures indicate that search construction is genuinely difficult and that errors are largely invisible to the person who made them.
A journal said my search is inadequate. What should I do?
Map the criticism onto the six PRESS elements, since that comment almost always points to one of them. Obtain a formal PRESS review if you have not had one, re-run the corrected strategy, and document every change. Report the full strategy for every database with dates and counts per PRISMA-S, since thin reporting is a frequent trigger for criticism. Present the revision as a documented, standards-based fix rather than an ad hoc patch, which is far more persuasive to an editor.
You May Also Find Useful
Validating the Stage You Cannot Redo
The search is the one part of a systematic review where an error cannot be undone after the fact, which is exactly why it is the part that most deserves an independent check before you commit to it. Build your primary strategy, have an information specialist review it against the six PRESS elements before you run it, correct what they find, and report the whole thing to PRISMA-S when you write up. Do that, and the search stops being the hidden weakness a peer reviewer exposes months later and becomes the part of your review that is demonstrably sound from the start.
If you want an information specialist trained in PRESS to review your search strategy before you run it, send us your question and draft strategy. You will have an itemized quote within 2 to 4 business hours, with no obligation.
