The focused question
The example below will guide you through the process of building comprehensive search strategy in PubMed database. PICO framework is used to formulate the focused question to help you build an effective search strategy to find relevant studies that meet the eligibility criteria for your review topic.
Does the use of light therapy help to improve sleep disturbances for dementia patients?
|
P(Population/ Patient) |
I(Intervention) |
C(Comparator) |
O(Outcome) |
|
Dementia / Alzheimer's disease |
light therapy
|
Usual care / placebo light |
|
According to the PRISMA 2020 Statement (published in 2021), details of search strategy should be clearly documented. We have designed the Search Template for you to plan your search, document the sources that you have searched, the search date and the search terms used (include both the controlled vocabulary and the keywords).
Please click on the link to download the template below. Follow the steps to fill out the question, the keywords and controlled vocabularies and your search strategy. Attach it in the email when you request for consultation.
Once you have identified the most important PICO elements of your research question (i.e. the intervention and patient terms), the next step is to brainstorm for keywords to come up with an extensive list of vocabulary to fill in the search template shown below. The more relevant and appropriate keywords are, the more comprehensive the search strategy will be.
For systematic review, the same keywords can be used to replicate the search across all databases.
|
Concept 1 /Population/Problem |
Concept2 /Intervention/Exposure |
|
|
Key concepts Identify the key concepts based on your research topic. |
Patients with dementia |
Light therapy |
|
Free text terms / natural language terms (synonyms, UK/US terminology, medical/laymen’s terms, acronyms/abbreviations, drug brands, more narrow search terms) List down your keywords for each concept. |
Dementia Dementias Alzheimer Alzheimer’s Alzheimers Lewy Body Lewy bodies |
Light therapy Light therapies Phototherapy Phototherapies Bright light therapy (BLT) |
After generating the keywords, the next step is to identify the subject headings or the controlled vocabulary or indexing terms.
What is Subject headings
Subject headings are the list of standardized vocabulary used by the indexer to describe the content of the articles in the databases. It is used interchangeably with controlled vocabulary or indexing term. Every database has their unique naming, e.g. in PubMed, it is known as MeSH (Medical Subject Heading).
Subject Hierarchy
You can make use of the MeSH Database in PubMed to search for the appropriate MESH term. This is the most precise way to find all the MEDLINE records via PubMed on a particular concept. Once the correct MeSH term is identified, click on the subject link to look out for the subject hierarchy shown below:
|
Dementia Primary Progressive Nonfluent Aphasia Diffuse Neurofibrillary Tangles with Calcification Frontotemporal Lobar Degeneration |
The subject terms are arranged in hierarchical order to show the broader and narrower terms. This is known as subject hierarchy. E.g. ‘Dementia’ is a broader term with the narrower terms indented underneath. When you search ‘Dementia’ as a MeSh term, all studies on dementia, include both general and specific will be found. Unlike subject heading, keyword search will only return you articles in which the word is used. You will miss studies which may use different terminologies to describe the same concept. Nevertheless, keyword complements subject headings as it retrieves the latest studies which will usually be missed as it takes time for indexer to index the articles.
For systematic review, both the MeSH and keywords should be combined to build a comprehensive search.
Based on the list of MeSH terms and the keywords identified, the next step is to build the search strategy as follows:
1.Search strategy for ‘P’ term (i.e. dementia)
|
P (Population) |
Building Search Statement for ‘P’ |
|
MeSH |
"Dementia"[Mesh] |
|
Keywords |
Dementia OR Dementias OR Alzheimer OR Alzheimer’s OR Alzheimers OR Lewy Body OR Lewy bodies |
|
Search Keywords in [Title/Abstract] |
Dementia[Title/Abstract] OR Dementias[Title/Abstract] OR Alzheimer[Title/Abstract] OR Alzheimer’s[Title/Abstract] OR Alzheimers[Title/Abstract] OR Lewy Body[Title/Abstract] OR Lewy bodies[Title/Abstract] |
Combine MeSH and keywords within the same concept with Boolean Operator ‘OR’
"Dementia"[Mesh] OR Dementia[Title/Abstract] OR Dementias[Title/Abstract] OR Alzheimer[Title/Abstract] OR Alzheimer’s[Title/Abstract] OR Alzheimers[Title/Abstract] OR Lewy Body[Title/Abstract] OR Lewy bodies[Title/Abstract]
2.Search strategy for ‘I’ term (i.e. light therapy)
|
I (Intervention) |
Building Search Statement for ‘I’ |
|
MeSH |
"Phototherapy"[Mesh] |
|
Keywords |
Light therapy OR Light therapies OR Phototherapy OR Phototherapies OR BLT |
|
Search Keywords in [Title/Abstract] field |
Light therapy[Title/Abstract] OR Light therapies[Title/Abstract] OR Phototherapy[Title/Abstract] OR Phototherapies[Title/Abstract] OR BLT[Title/Abstract] |
Combine MeSH and keywords within the same concept with Boolean Operator ‘OR’
"Phototherapy"[Mesh] OR Light therapy[Title/Abstract] OR Light therapies[Title/Abstract] OR Phototherapy[Title/Abstract] OR Phototherapies[Title/Abstract] OR BLT[Title/Abstract]
3.Combining searches for ‘P’ AND ‘I’ with Boolean Operator ‘AND’
|
("Dementia"[Mesh] OR Dementia[Title/Abstract] OR Dementias[Title/Abstract] OR Alzheimer[Title/Abstract] OR Alzheimer’s[Title/Abstract] OR Alzheimers[Title/Abstract] OR Lewy Body[Title/Abstract] OR Lewy bodies[Title/Abstract]) AND ("Phototherapy"[Mesh] OR Light therapy[Title/Abstract] OR Light therapies[Title/Abstract] OR Phototherapy[Title/Abstract] OR Phototherapies[Title/Abstract] OR BLT[Title/Abstract]) Results retrieved = 382 (search last updated on 15 Aug 2023) |
4.Refining the final search
If there is a large result set retrieved, the search strategy can be further refined depending on the type of questions being asked. Filters are available in PubMed and most of the databases to help narrow down the search results. Please refer to the Search Filters in the next section for more details.
Additionally, you should also refer to the PRESS Evidence-Based Checklist for practical tips to improve the quality of your search strategy.
McGowan, J., Sampson, M., Salzwedel, D. M., Cogo, E., Foerster, V., & Lefebvre, C. (2016). PRESS Peer Review of Electronic Search Strategies: 2015 Guideline Statement. J Clin Epidemiol, 75, 40-46. https://doi.org/10.1016/j.jclinepi.2016.01.021
In addition to the Boolean operators, most of the subscribed databases allow searching using proximity operators which is lacking in the free databases, e.g. PubMed.
Proximity operators is a technique use to locate the occurrence of search words that are closed by one another. It is usually denoted by NEAR/n (n is the number of words) to find the search terms within a specific number of words of each other. The proximity symbol may vary depends on the databases used. Please refer to the help guide of the respective databases to look out for the correct syntax.
Proximity operators provide an alternative to ‘AND’ operator. For example, searching (dementia NEAR/6 “sleep disorder”) will retrieve more relevant set of results than searching (dementia AND “sleep disorder”). The former shows that ‘dementia’ and “sleep disorder” appear 6 or less words within one another. It is more precise than the latter which is not able to show how far apart are the words from one another.
Please find below a concised summary of the proximity operators of the commonly used databases. For more details, please refer to the NUS Libraries Database Syntax Guide below.
|
Databases |
Discipline |
Controlled vocabulary |
Proximity Operators |
|
MEDLINE(PubMed) |
Biomedical and Life Sciences |
MeSH |
"search terms"[field:~N] search terms= Two or more words enclosed in double quotes field = The search field tag for [Title] or [Title/Abstract] fields. N = The maximum number of words appearing between your search terms |
|
Embase |
Biomedical and Life Sciences |
Emtree |
NEAR/n (in any order) up to (n-1) words in between NEXT/n (in same order) up to (n-1) words in between |
|
The Cochrane Library |
Medical & Health Sciences |
MeSH |
NEAR/n (in any order) up to (n-1) words in between NEXT/n (in same order) up to (n-1) words in between |
|
CINAHL(Ebscohost) |
Nursing and allied health |
CINAHL Subject Headings |
Nx: (in any order) up to (x) words in between Wx: (in same order) up to (x) words in between |
|
PsycINFO(Ovid) |
psychology and the behavioral and social sciences |
Thesaurus of Psychological Index Terms |
ADJ finds two terms next to each other in the specified order ADJn (in any order) up to (n-1) words in between |
|
ERIC(ProQuest) |
Education |
ERIC Thesaurus |
NEAR/n or N/n (in any order) up to (n) words in between PRE/n: (in same order) up to (n) words in between |
|
Scopus |
Multidisciplinary |
NIL |
W/n (in any order) up to (n) words in between PRE/n (in same order) up to (n) words in between |
|
Web of Science |
Multidisciplinary |
NIL |
NEAR/x (in any order) up to (x) words in between |
A compilation of search syntax (e.g. search operators, wildcards, phrase searching, search field, etc) of selected databases at your fingertips for your easy reference when conducting a comprehensive search for systematic review.