BETs Search Strategies

What is a search filter?

A methodological search filter is defined as:

"A search term or terms (such as 'random allocation' for sound studies of medical intervention) that select studies that are at the most advanced stages of testing for clinical application." (Wilczynski et al 1995, p436)

In other words, it is a predefined search strategy designed to retrieve levels of evidence (RCTs, systematic reviews etc) or types of clinical queries (diagnosis, prognosis, etiology, treatment) when combined with the subject search terms of your choice. They are also referred to as hedges, Clinical Queries (USA), or optimal search strategies.

Called 'methodological' search strategies since the strategies are made up of search terms that relate to the methodology of study design. For example, an RCT filter may contain terms such as double blinded, randomized, clinical trial. A diagnosis filter may contain terms such as sensitivity, specificity, etc.

Main Aims of a Search Filter

Designed to overcome imprecise search skills of end-users in the MEDLINE database they also compensate for indexing inconsistencies inherent in the Medline database.

Methodological search filters were first devised by Haynes et al, at McMaster University in 1994. Search terms relating to diagnosis, prognosis, therapy and etiology were collated and ran through a computer program to identify optimal search strategies which were then tested and validated in the Medline database and compared against a 'gold standard' of known articles in each of the 4 areas. Further studies have tested and validated search filters in a number of areas relating to research methodology and evidence-based medicine. (Dickersin 1994, White 2001).

Search filters are NOT a guarantee of retrieving quality research. While they will retrieve studies relating to research methodology the onus remains on the user to appraise the evidence for QUALITY, RELEVANCE and APPLICABILITY.

Types of Filters Available

Filters currently available are system specific (OVID, Dialog, PubMed, etc.) and database specific (Medline, Cinahl, Embase, PsychInfo) covering a range of publication types (RCT, systematic reviews, diagnosis, etc.). Depending on user needs filters are designed to be highly sensitive, precise or one-line strategies. For BETs searches we recommend the highly sensitive filters be used to ensure no evidence is overlooked.

Example (and explanation of commands)

Methodological search filters are made up of MeSH and freetext search terms and use advanced search operators (exp, .pt., adj, etc).

RCT filter for OVID (Dickersin K et al, 1996)

RANDOMIZED CONTROLLED TRIAL.pt.
CONTROLLED CLINICAL TRIAL.pt.
RANDOMIZED CONTROLLED TRIALS.sh.
RANDOM ALLOCATION.sh.
DOUBLE BLIND METHOD.sh.
SINGLE-BLIND METHOD.sh.
or/1-6
ANIMALS.sh. not HUMAN.sh.
7 not 8
CLINICAL TRIAL.pt.
exp CLINICAL TRIALS
(clin$ adj25 trial$).ti,ab.
((singl$ or doubl$ or trebl$ or tripl$) adj25 (blind$ or mask$)).ti,ab.
PLACEBOS.sh.
placebo$.ti,ab.
random$.ti,ab.
RESEARCH DESIGN.sh.
or/10-17
18 not 8
19 not 9
COMPARATIVE STUDY.sh.
exp EVALUATION STUDIES
FOLLOW UP STUDIES.sh.
PROSPECTIVE STUDIES.sh.
(control$ or prospectiv$ or volunteer$).ti,ab.
or/21-25
26 not 8
27 not (9 or 20)
9 or 20 or 28

Commands

Command		Function
/	Subject heading (MeSH)	Retrieves all items with the subject heading assigned to it by the indexers
.exp.	Explode	Retrieves all items with the subject heading and all associated narrower MeSH terms assigned
.ti.	Title	Retrieves items with words contained in the title
.ab.	Abstract	Retrieves items with words contained in the abstract
.mp.	Keyword	Retrieves items with words contained in the title, abstract or subject heading
.pt.	Publication type	Retrieves items of a specified publication type (e.g. Multicenter study)
$	Truncation	Replaces any number of characters including zero (e.g. child$ will retrieve items containing child, children, childhood, etc.)
adj	Adjacency operator	Retrieves items with query terms on either side in the specified order
N.B. Other systems may use different commands, refer to the 'help' page in the particular system/database in which the search is conducted for further explanations.

Why use Search Filters

Search filters are a useful way of limiting broad searches. They apply a level of consistency to your searching making it more systematic. Above all they can save you time in conducting searches since tested and validated search strategies exist which present a list of search terms which are known to be effective in retrieving particular types of study.

When to use Search Filters

Used to limit large search results or when looking for a particular publication type (RCTs, etc) or clinical query (diagnosis, etc).

How to use Search Filters

Filters are an advanced searching technique and may seem daunting at first but don't be put off by them, as they can be an extremely useful tool in retrieving quality research from the medical literature. To use them follow the instructions below.

Conduct a highly sensitive subject search combining MeSH and free text.
Check that the filter is relevant to the system, database and study design.
To avoid typographical errors highlight each line in turn, copy and paste each line of the search into the database and perform search. Alternatively type each line in manually. (Take care to note any changes in the written numbers of search statements when combining with 'OR'/'AND' e.g. 'or/10-17', which may occur as a result of adding on the search filter to your subject search. To avoid confusion it may be easier to type in the filter first and save it, then start a new subject search and re-run the saved filter at the end).
Use 'AND' to combine the results of the subject search with the filter.

Although filters may be downloaded to disk, at this time they cannot be uploaded from disk into OVID and therefore they have to be typed in manually. For frequent users, type in the filter and save it as a permanent search. The filter can then be re-run as required. If in doubt seek help from a librarian/information professional.

Things to Consider When Using Filters

Search filters are NOT a guarantee of retrieving 'quality' research'. You still need to critically appraise results for QUALITY and RELEVANCE.
A filter developed in Medline will not have the same sensitivity/specificity when run in Embase for example, since the indexing language used in the two databases is different. It is important to ensure that the search filter is compatible with the database and the system (Ovid, PubMed etc.).
Due to annual amendments and updates to MeSH the search terms in the filter should be checked annually to see if there are any new terms which may be relevant.
Not relevant to all research areas (eg public health) or were the research base is small.
Not all filters published on the www have been tested and validated.
The aim of your search is to locate the 'best evidence'. It is important to note that specific types of filter (e.g. RCT) will not neccessarily pick up all systematic reviews or meta-analyses. Hence it is recommended that a search using filters also aims to retrieve such reviews.

Finally, there will always be some articles missed and erroneous ones captured, however search filters have been shown to improve upon end-user searching (Dickersin 94).

Filters on the www

Filter Type	Available At
RCT	Cochrane Library (can be found under Appendices>Appendix 5b)
Systematic reviews Meta-analysis	NHS centre for reviews and dissemination
Diagnosis Etiology	CASPfew
Therapy Prognosis	PubMed

BestBETs subject filters in emergency medicine

Emergency physicians face two additional problems in their search for the evidence. Firstly, Emergency medicine is a broad specialty covering many topic areas and secondly, evidence of the highest quality is not always available in emergency medicine (Mackway-Jones et al 1998). Therefore, the subject becomes the focus of the search. Initially searches need to be highly sensitive to ensure important evidence is not overlooked.

Updated Paediatric filter incoporating MeSH 2003 changes

BestBETs Paediatric Filter 2003 (PDF - 108KB)

Please note that the filter is a prototype and as such is still in the testing stages, further amendments to the filter may result from more rigourous testing.

Poster presented at the 8th European Conference of Medical and Health Libraries, Cologne, 16-21 September 2002.

BestBETs Paediatric Filter (PDF - 1.22MB)

Copyright: The poster is the property of www.bestbets.org. No part of the poster may be reproduced without prior permission of the authors.

Information officers at BestBETs are currently devising further subject filters to be used in the BETs search strategy. These filters will appear here once testing is complete.

References

Haynes RB, Wilczynski N, McKibbon KA et al. Designing optimal search strategies for detecting clinically sound studies in Medline. Journal of the American Medical Informatics Association 1994;1:447-59.
Dickersin K, Scherer R, Lefebvre C. Systematic reviews: Identifying relevant studies for systematic reviews BMJ 1994,309(6964),1286-91.
Dickersin K, Larson K. Establishing and maintaining an international register of RCTs. In: The Cochrane Library. Oxford: Update Software, 1996.
Mackway-Jones K, Carley SD, Morton RJ et al. The best evidence topic report: A modified CAT for summarising the available evidence in emergency medicine. Journal of Accident and Emergency Medicine 1998;15:222-26.
White VJ, Glanville JM, Lefebvre C et al. A statistical approach to designing search filters to find systematic reviews: objectivity enhances accuracy. Journal of Information Science 2001;27(6):357-70.
Wilczynski NL, Walker CJ, McKibbon KA et al. Reasons for the loss of sensitivity and specificity of methodologic MeSH terms and textwords in MEDLINE. AMIA 1995;436-40.