Professional Documents
Culture Documents
Medical
databases
Rhanderson Cardoso, MD
Cochrane
Cochrane is a British institution with the mission of conducting and organizing scientific
research in health fields, especially in systematic reviews and meta-analyses.
We will typically include Cochrane as one of the search engines for our systematic
reviews and meta-analyses.
Cochrane is also the main authority on how to conduct systematic reviews and meta-
analyses. We will use their references anytime a question comes up on how to do a given
step in the process.
Cochrane Handbook for Systematic Reviews of Interventions:
https://training.cochrane.org/handbook/current
Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy:
https://training.cochrane.org/handbook-diagnostic-test-accuracy
Web of Science
This database is developed by Clarivate, an analytics company with multiple products in
science and business.
Clarivate is also responsible for calculating the famous “impact factor” of journals.
It is a collection of several international databases, for example:
Chinese Science Citation Database;
SciELO, Latin America, Caribbean, and Iberian countries’ database;
Arabic Regional Citation Index.
Web of Science requires a subscription for access. Again, if you want to use Web of
Science in your search (you don’t have to), check whether you have institutional access.
ClinicalTrials.gov
This is a registry of clinical trials, with the goal of improving access of the public to
clinical trials.
In the USA, it is mandatory to register clinical trials prospectively in this database. The
registry will include important protocol information, such as the target population,
study arms, planned interventions, etc.
This registration is mainly used to inform the study design and stage of development,
which is typically updated throughout the conduction of the study.
When the study is finalized, institutions and authors responsible for the conduction of
the study usually publish the results in a journal (indexed in PubMed and other
databases) before the results are updated on ClinicalTrials.gov.
Therefore, it is rare that we need to get the results of studies from ClinicalTrials.gov
since most of the results of clinical trials will be published in the literature, far before
they are updated in ClinicalTrials.gov.
However, theoretically, it’s possible that authors or sponsoring companies in charge of a
given clinical trial decide not to publish the results or delay the publication of results.
This may happen, for example, if the company has no intention of continuing the
development of a drug.
In these cases, there may still be a regulatory need to publish the results on
ClinicalTrials.gov, so you may only find some results there.
Gray literature
Asides from the previously mentioned databases, for some specific topics, it’s important
to search the gray literature.
Gray literature includes everything that is produced by government agencies, academic
fields, businesses, or industries that are not controlled by publishing companies.
For instance, if you are writing a systematic review of public policies on the
management of dengue fever in Latin America, there will probably be little information
about it on PubMed. So, it is important to search the websites of public health agencies,
send them emails, etc. All of this is considered gray literature.
Attention: for most systematic reviews and meta-analyses, there is no need to search
the gray literature. It is only if you search for some specific topic that may not be
available in indexed journals that you should look in the gray literature.
Importantly, there is no specific protocol on how the gray literature search must be
conducted. Each topic will have a different protocol of how to search for this data.
Protocol registration
QUESTIONS
1. Among the following search engines, which ones must be included when the research question
involves topics that are particular to Latin America?
a) LILACS
b) Embase
c) Web of Science
d) ClinicalTrials.gov
2. The advantage of including Scopus or Embase as one of the medical databases in your search is:
a) Scopus and Embase are the only databases with controlled vocabulary.
b) Scopus and Embase include only high-impact journals.
c) Scopus and Embase include conference abstracts.
d) Scopus and Embase are the only public-access platforms.
5. Which alternative is correct about search engines for different medical databases?
a) The search strategy may be modified to fit different search engines, but this may not be needed in
most cases.
b) The search strategy must be the exact same in all search engines.
c) The search strategy must be specific for each medical database, with the use of field tags that are
appropriate for each database.
d) Controlled vocabulary must be used for search strategies in PubMed (MeSH) and Embase (Emtree).
Exercise 1.
A. LILACS is the largest search engine for Latin-American studies. Consider including LILACS as a
search tool in studies with diseases of higher prevalence in Latin America.
Exercise 2.
C. Scopus and Embase are Elsevier platforms. They are not open access, but they have a major
advantage – they include conference abstracts. If an abstract has the data you need, it may be
included in your systematic review and meta-analysis. If you have many studies that are published
already, you may set conference abstracts as an exclusion criterion in your meta-analysis. It is up to
you.
Exercise 3.
B. The protocol registration has been increasingly requested by journals before publishing a meta-
analysis. The registration must be done before conducting the search and analyzing the results. This is
the reason why we put it at this stage in the Meta-Analysis Academy training program.
Exercise 4.
D. The gray literature must be searched for specific topics that are not typically found in indexed
databases. It is a generic term to represent everything that is produced by government agencies,
academic fields, businesses, or industries, but is not controlled by commercial publishing companies.
Consider using it in projects that include public health policies, for example. There is no universal
protocol on how the gray literature must be searched. It will depend entirely on the topic of the
search.
Exercise 5.
A. The search strategy may be modified for different search engines, but it does not have to. If you
build a simple, effective search strategy with synonyms, Boolean operators, parentheses, and
quotation marks, without anything specific to a given database, you may copy and paste it into
different search engines for many databases.