Skip to Main Content
It looks like you're using Internet Explorer 11 or older. This website works best with modern browsers such as the latest versions of Chrome, Firefox, Safari, and Edge. If you continue with this browser, you may see unexpected results.

Systematic reviews: Searching effectively

A quick guide to introduce you to systematic reviews

Search techniques

It is possible to use a number of different keywords or search terms in a single search, by using operators

Operators link your search terms and define the relationship between them. This enables more accurate searching and therefore more relevant results being returned. It also saves you time as you don't have to carry out numerous similar searches where just one or two search terms are changed each time.

 

AND

Narrows down a search and works by looking for both terms you specify (in any order). This narrows a search as both terms must appear. 

e.g. mindfulness AND chronic pain

OR

Broadens a search and works by looking for either of the terms you specify, and will find articles where either term or both terms are mentioned (in any order). This broadens a search as is useful for retrieving synonyms or variant spellings. 

e.g. depression OR anxiety

NOT

Excludes terms from the search. This operator isn't recommended for systematic reviews as it is highly likely you will lose relevant studies. You will need to manually filter out unwanted results. 

e.g. diabetes NOT gestational‚Äč

 

You should use database filters and limits with caution when conducting systematic reviews due to the risk of introducing bias. If you want to limit your results to a particular study type, or exclude animal studies from your results, you should consider using a sensitive search filter that has been created by expert searchers such as Information Specialists.

Search filters are pre-designed sensitive search strategies that can be used to identify a particular type of record. They are an effective way of locating particular types of studies (i.e. RCTs or observational studies). 

They can be copied and pasted into the search box to run a search for a particular type of study. Searchers can combine the results of the search filter with their result set using the Boolean operator AND to filter a particular study type from their results.

SIGN filters have been designed to be used in the databases Medline, Embase and CINAHL, hosted on the Ovid platform. 

Cochrane search filters are available for the databases MEDLINE and EMBASE.

CADTH ( Canadian Agency for Drugs and Technologies in Health) filters include those to help you find economic evaluations, quality of life studies, RCTs, guidlines and systematic reviews.

  • Human studies filter

The following Cochrane designed search filter can be used in Ovid Medline to locate animal studies. You can then use remove these from your search results using the NOT operator.

animals.sh. not (humans.sh. and animals.sh.)

To see an example of this, take a look at this demonstration search strategy from the Cochrane Handbook on Tamoxifen for breast cancer. The human studies filter has been used in line 10, in conjunction with the publication type filter used in lines 1-9. You will see in line 11 that the animal studies identified by this filter have been removed from the results of the publication type filter search results (9 NOT 10).

Search filters designed by The InterTASC Information Specialists' Sub-Group (ISSG)

Using a combination of keyword and subject heading terms in your search strategy can be an effective way of ensuring that your search retrieves a greater number of relevant results. Indexers allocate subject headings to studies which allow you to search at subject level rather than just keyword level. This way of searching can be helpful as different authors may use different terminology for the same concepts and these references may be missed from your keyword searches alone.

Subject headings are provided in databases such as MEDLINE, EMBASE and APA PsycInfo, although each database has a different approach to indexing and many use different headings. For this reason, it is strongly recommended that you modify your search strategy for each database (incorporating the most relevant subject heading) and run each database search separately.

Watch the video below for a demonstration of how to carry out an advanced search using keywords and subject headings on Ovid databases.

Tip: To help you locate subject headings for your search strategy, it is recommended that locate articles of relevance in the database you are using and scan the full database records for keywords and make note of the subject headings allocated to these articles. You can then include these in your search strategy to retrieve other relevant studies.

The Cochrane Handbook defines sensitivity and precision as follows:

"Sensitivity is defined as the number of relevant reports identified divided by the total number of relevant reports in existence. Precision is defined as the number of relevant reports identified divided by the total number of reports identified."

JPT, Green S (editors). 6.4.4 Sensitivity versus precision, Cochrane Handbook of Systematic Reviews of Interventions. Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.handbook.cochrane.org.

Searches carried out for systematic reviews should aim for high sensitivity, which may lead to low precision. You may find that your systematic search strategy retrieves a large number of results which you will need to screen for relevance and inclusion.

Proximity searching: Using an operator such as ADJ or NEAR  to define the proximity between your keywords can be helpful in increasing the precision of your results compared with using AND to combine the keywords e.g. gestational adj3 diabetes (returns results with the keywords "gestational" and "diabetes" within three words of each other in either direction. Tip - check the help section of the database you are using as these operators differ across different databases.

Proximity searching should be used with caution in systematic reviews to avoid losing relevant studies.

Contact Us