Once the final search has been done, combining all concepts with AND, you may refine the search results further, e.g. by publication date, study design.
It is generally not advisable to use the limit options that the databases provide:
TIP! It is a better option to make the desired refinement part of your search strategy (e.g. only find human studies or only find RCTs). For more information see Methodological search filters below.
Methodological search filters are search terms/strategies to identify a topic or aspect (e.g. study type, age group). They are "tried and tested" strategies intended for repeated use. If there is a relevant study type filter available, add it to your search.
Sources of filters are:
Useful articles for observational studies:
Limiting study inclusion on the basis of language of publication introduces language bias. Both Cochrane and the Campbell Collaboration advise not to restrict the searches by language. JBI recently acknowledged that limiting searches to English is common practice in JBI systematic reviews but that this should be reconsidered.
For the example, in Step 1, you may have decided to only include studies that are randomised controlled trials (RCTs). For a systematic review it is not recommended to use the limit options that the databases provide. It is a better option to make this refinement part of the search strategy. According to the Cochrane Handbook, you may like to add the Highly Sensitive Search Strategy filter for identifying RCTs to the search. It is added at the end of the search strategy.
This is the Medline via Ovid format for the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity-maximizing version (2008 revision):
1 randomized controlled trial.pt.
2 controlled clinical trial.pt.
3 randomized.ab.
4 placebo.ab.
5 drug therapy.fs.
6 randomly.ab.
7 trial.ab.
8 groups.ab.
9 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8
10 exp animals/ not humans.sh.
11 9 not 10
From Cochrane Handbook, Technical Supplement to Chapter 4, p61, Box 3.c
You can run this search strategy after the final search, which combines all four concepts with AND (search 44 in Medline via Ovid Example)
Note! This filter outlines the proper exclusion of animal studies in a search (see double negative in line 10 and 11)
This is the PubMed format for the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity-maximizing version (2008 revision):
#1 randomized controlled trial [pt]
#2 controlled clinical trial [pt]
#3 randomized [tiab]
#4 placebo [tiab]
#5 drug therapy [sh]
#6 randomly [tiab]
#7 trial [tiab]
#8 groups [tiab]
#9 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8
#10 animals [mh] NOT humans [mh]
#11 #9 NOT #10
From Cochrane Handbook, Technical Supplement to Chapter 4, p59-60, Box 3.a
You can run this search strategy after the final search, which combines all four concepts with AND (search #49 in Medline via PubMed Example).
Note! This filter outlines the proper exclusion of animal studies in a search (see double negative in line #10 and #11)
If you would like to limit your search to Human studies only, please do not use the Human filters provided by the databases. It is best practice to identify animal-only studies first, and then exclude those from the search results (double negative). See a selection of search strategies below:
Medline via Ovid
from Cochrane Handbook, Technical Supplement to Chapter 4 (p62):
not (exp animals/ not humans.sh.)
Medline via PubMed
from Cochrane Handbook, Technical Supplement to Chapter 4 (p61):
NOT (animals [mh] NOT humans [mh])
Embase via Ovid
from Embase: Excerpta Medica Database Guide - Limits: Humans only (removes records about animals):
not ((exp animal/ or exp invertebrate/ or nonhuman/ or animal experiment/ or animal tissue/ or animal model/ or exp plant/ or exp fungus/) not (exp human/ or human tissue/))
PsycINFO
from https://doi.org/10.1177/00236772211045485
NOT (filter described in Supplemental material 5)
CINAHL via EBSCOhost
from Cochrane Handbook, Technical Supplement to Chapter 4 (p66):
NOT (((MH "Animals+") OR (MH "Animal Studies") OR (TI "animal model*")) NOT (MH "human"))
Scopus
Scopus includes MeSH and Emtree terms from Medline and Embase records so you could adapt Medline's and Embase's Limit to Humans strategy:
AND NOT ((INDEXTERMS(animals OR animal)) AND NOT (INDEXTERMS(humans OR human)))
Web of Science
from https://doi.org/10.1177/00236772211045485
NOT (filter described in Supplemental material 6)
