Developing a Searchable Review Question

Overview

A searchable review question connects the knowledge need to a reproducible evidence-identification plan. It states what evidence is relevant, while a search strategy translates selected parts of that question into the syntax and vocabulary of one database. These are related but different products.

A good review question is precise enough to support eligibility criteria and synthesis, yet broad enough to capture the relevant evidence. Overly broad questions retrieve an unmanageable and conceptually mixed literature. Overly narrow questions may miss relevant studies because authors use different terminology or because the question has imposed details that are not essential.

Why this matters

Search quality cannot compensate for a poorly framed question. When review boundaries are vague, the team may change eligibility decisions during screening, include studies that answer different questions or exclude inconvenient evidence after results are known. Conversely, putting every element of a detailed protocol into the database query can make the search so restrictive that relevant studies are lost.

The question should define relevance. The search should retrieve records likely to contain relevant evidence. Because titles, abstracts and indexing are incomplete, the search often represents only the most searchable core concepts rather than every eligibility criterion.

Question first, search second

A review question should identify:

• the population, setting or evidence domain;
• the exposure, intervention, phenomenon or concept of interest;
• any essential comparator;
• the outcome, experience or evidence characteristic of interest;
• the eligible study or source types; and
• the intended synthesis or decision.

The database search statement is then built from concepts that are consistently described in searchable fields [1]. Study design, outcome or setting terms may be omitted from the query when they are poorly indexed or likely to reduce sensitivity.

Choosing a question framework

Frameworks are prompts, not rigid templates. Choose one that represents the review purpose.

PICO for interventions

PICO commonly represents Population, Intervention, Comparator and Outcome. It is well suited to focused questions about intervention effects [2]. The comparator and outcome may be essential to eligibility but not always useful as search blocks.

Example: Among women with recent gestational diabetes, do reminder-based follow-up interventions, compared with usual care, increase completion of postpartum glucose testing?

PECO for exposures or aetiology

PECO uses Population, Exposure, Comparator and Outcome. It can structure questions about environmental, behavioural or clinical exposures.

Example: Among women with gestational diabetes, is greater travel distance to follow-up services, compared with shorter distance, associated with non-completion of postpartum testing?

PCC for scoping reviews

PCC uses Population, Concept and Context. It supports broad mapping questions where a narrow comparator or outcome is not central [3].

Example: What patient, provider and health-system barriers to postpartum diabetes follow-up have been reported among women after gestational diabetes across maternity and primary-care settings?

PICo and SPIDER for qualitative evidence

PICo can represent Population, Phenomenon of Interest and Context. SPIDER represents Sample, Phenomenon of Interest, Design, Evaluation and Research type [4]. These frameworks can help articulate qualitative evidence needs, but design and research-type filters should be used carefully because qualitative studies are inconsistently indexed.

Example: How do women with recent gestational diabetes experience the transition from maternity care to postpartum metabolic follow-up?

Other frameworks

Diagnostic, prognostic, prevalence, policy and implementation reviews may require modified frameworks. The important principle is that the structure should reflect the intended inference, not that every question must be forced into PICO [5].

Alternative structures for one health topic

illustrative

Table 1. Review questions derived from postpartum follow-up

Purpose	Framework	Structured question
Intervention effect	PICO	Do reminders versus usual care increase test completion among women after gestational diabetes?
Association	PECO	Is travel burden associated with non-completion of postpartum testing?
Evidence mapping	PCC	What barriers, interventions and measurement approaches have been reported across care settings?
Experience	PICo	How do women experience referral and follow-up after gestational diabetes?

Note: These questions may share search concepts but require different eligibility and synthesis plans.

Extracting searchable concepts

A search is usually organised into concept blocks. Synonyms within a block are combined with OR; different concept blocks are combined with AND.

For the intervention-effect question, possible core blocks are:

1. gestational diabetes;
2. postpartum or postnatal period; and
3. follow-up testing or screening.

A reminder-intervention block might be included if the review is specifically restricted to reminders. An outcome term such as “completion rate” may be omitted because relevant records may not mention completion in the title or abstract.

DECISION GUIDE

Should this element become a search block?

Include an element when it is:

• central to relevance;
• described with reasonably identifiable terminology;
• likely to appear in searchable fields or indexing; and
• unlikely to remove many eligible records.

Consider screening rather than searching for an element when terminology is inconsistent, reporting is poor or the criterion is too detailed for reliable retrieval.

Building a concept table

For each concept, identify:

• preferred clinical and public-health terms;
• abbreviations and full forms;
• older terminology;
• spelling variants;
• singular, plural and word-stem variants;
• drug, intervention or programme names;
• condition subtypes;
• controlled-vocabulary headings; and
• terms found in known relevant papers.

Concept table for postpartum glucose follow-up

illustrative

Table 2. Initial search-concept table

Concept	Free-text examples	Controlled-vocabulary direction
Gestational diabetes	gestational diabetes, diabetes in pregnancy, GDM	Database heading for diabetes, gestational
Postpartum period	postpartum, post-partum, postnatal, after delivery, puerperium	Database heading for postpartum period
Glucose follow-up	glucose test, oral glucose tolerance test, diabetes screening, metabolic follow-up	Headings for glucose tolerance test, mass screening or follow-up studies as appropriate

Note: Controlled vocabulary and syntax must be checked separately in each database.

Known relevant studies are useful for discovering vocabulary. Examine their titles, abstracts, author keywords and subject headings. However, do not design the search so narrowly that it retrieves only those known records.

Controlled vocabulary and free text

Bibliographic databases may assign standard subject headings, such as Medical Subject Headings (MeSH) in MEDLINE. Controlled vocabulary groups records that use different author language under a common indexed concept. Free-text searching captures recent records not yet indexed, emerging terminology, wording not represented by a heading and concepts indexed inconsistently.

Robust searches commonly combine both. A subject heading in one database cannot be copied unchanged into another database because thesauri, field codes and explosion rules differ.

Setting review boundaries

Important boundaries include:

• population characteristics;
• setting and geographic scope;
• intervention or exposure definitions;
• eligible outcomes or phenomena;
• study designs or source types;
• publication period;
• language;
• publication status; and
• minimum follow-up or data requirements.

Each restriction should be linked to the review purpose. Date restrictions may be justified by a policy change, diagnostic technology, treatment era or prior high-quality review. Language restrictions may be unavoidable, but they should be reported as a limitation rather than treated as methodologically neutral.

PROTOCOL CHECKLIST

Record decisions before running the final search

• State the primary review question and intended synthesis.
• Define each eligibility domain operationally.
• Identify the core searchable concepts.
• Build and retain a concept-synonym table.
• Identify known relevant records for later search testing.
• Justify date, language, setting and publication-type restrictions.
• Decide which criteria will be applied during screening rather than searching.
• Record unresolved terms for consultation with a librarian or information specialist.

Testing the question and concepts

A preliminary search is an iterative design step. It can reveal whether terminology is too broad, whether important subgroups use different labels and whether the expected evidence exists. Search development may refine the wording of the review question, but substantive changes should be documented rather than silently made after seeing results.

Current evidence-retrieval guidance recommends iterative development, testing and full documentation of the search [6]. Testing should ask:

• Are known relevant records retrieved?
• Are key concept blocks causing unexpected loss?
• Which irrelevant meanings dominate the results?
• Are important terms absent from the concept table?
• Is the volume manageable without compromising the question?

Conceptual versus operational sensitivityMMed extension

A search can be technically complex yet conceptually incomplete. Search sensitivity depends not only on Boolean syntax but also on whether the review team has represented the concept’s clinical, social and historical language. Stakeholder consultation, thesaurus browsing and examination of diverse seed papers can expose terminology that one disciplinary team might miss.

Worked example

WORKED EXAMPLE

Refining the review question and search blocks

Initial question: What is known about postpartum follow-up after gestational diabetes?

Problem: The question mixes prevalence, determinants, experiences and interventions.

Selected review purpose: Map barriers and facilitators to completion of postpartum glucose testing.

Revised question using PCC: Among women following a pregnancy affected by gestational diabetes (Population), what barriers and facilitators to postpartum glucose testing (Concept) have been reported across maternity, primary-care and community settings (Context)?

Core search blocks: Gestational diabetes AND postpartum period AND glucose follow-up.

Criteria reserved for screening: The source must report a barrier or facilitator; testing must occur after delivery; and the source must provide empirical or clearly described programme evidence.

Why not search the words “barrier” and “facilitator” only? Relevant studies may report predictors, experiences, access problems or implementation determinants without using those exact labels.

Appraising a searchable question

APPRAISAL CHECKLIST

Review question and search-concept appraisal

• Does the question identify a coherent purpose and intended synthesis?
• Is the selected framework appropriate rather than mechanically applied?
• Are the population, concept and context boundaries explicit?
• Do the search blocks represent the essential concepts without encoding every eligibility criterion?
• Are both free-text and controlled-vocabulary terms planned?
• Have terminology and restrictions been tested against known relevant records?
• Are pragmatic restrictions justified and documented?

Suggested protocol wording

Common mistakes

Knowledge check

Why might the outcome be omitted from a PICO-based database search?

Answer and reasoning

Outcome terminology may be diverse or absent from titles, abstracts and indexing. Requiring the outcome can therefore remove eligible studies. The outcome can remain an essential eligibility criterion assessed during screening even when it is not a search block.

Which framework is usually more suitable for a broad scoping question about concepts and settings: PICO or PCC?

Answer and reasoning

PCC is usually more suitable because it prompts the Population, Concept and Context without forcing a comparator or narrowly specified outcome. The choice still depends on the actual purpose rather than the review label alone.

A search uses only the phrase ‘gestational diabetes’. What is missing?

Answer and reasoning

The team should assess spelling and wording variants, abbreviations, older terminology and relevant controlled-vocabulary headings. Depending on the database and period, relying on one phrase may miss records described or indexed differently.

Why should date restrictions be justified rather than used automatically?

Answer and reasoning

An arbitrary date restriction can exclude relevant evidence and distort conclusions. A defensible restriction should follow from the review purpose, such as a major policy change, technology era or the end date of a prior high-quality review.

Key take-home messages

SUMMARY

Key take-home messages

• The review question defines relevance; the search strategy retrieves records likely to contain relevant evidence.
• Eligibility criteria and search concepts overlap but are not identical.
• Use a question framework that matches the review purpose.
• Build concept blocks from free-text terms, variants and database-specific controlled vocabulary.
• Test and document boundaries before the final search.

References

1. Booth A. Clear and present questions: formulating questions for evidence based practice. Libr Hi Tech. 2006;24(3):355–368. https://doi.org/10.1108/07378830610692127
2. Richardson WS, Wilson MC, Nishikawa J, Hayward RSA. The well-built clinical question: a key to evidence-based decisions. ACP J Club. 1995;123(3):A12–A13.
3. Peters MDJ, Marnie C, Tricco AC, Pollock D, Munn Z, Alexander L, et al. Updated methodological guidance for the conduct of scoping reviews. JBI Evid Synth. 2020;18(10):2119–2126. https://doi.org/10.11124/JBIES-20-00167
4. Cooke A, Smith D, Booth A. Beyond PICO: the SPIDER tool for qualitative evidence synthesis. Qual Health Res. 2012;22(10):1435–1443. https://doi.org/10.1177/1049732312452938
5. Munn Z, Stern C, Aromataris E, Lockwood C, Jordan Z. What kind of systematic review should I conduct? A proposed typology and guidance for systematic reviewers in the medical and health sciences. BMC Med Res Methodol. 2018;18:5. https://doi.org/10.1186/s12874-017-0468-4
6. Lefebvre C, Glanville J, Briscoe S, Featherstone R, Littlewood A, Metzendorf M-I, et al. Technical Supplement to Chapter 4: Searching for and selecting studies. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. Version 6.5. Cochrane; 2024.