Systematic review search strings can be incredibly complex. This is why it is fundamental to have an information specialist on the research team. For an example of a systematic review search strategy, please examine the document below to see additional information from M.D. Anderson Cancer Center Library.
A preliminary search is about planning and developing your protocol. It helps the research team understand what the review would/should entail, and the preliminary search will also give an estimate for the time and funding commitment necessary for the investigation. Researchers should assume an exhaustive search will identify about 2-3 times more the number of citations than the preliminary search.
1. Searching for research trends
You need to see a sample of the research trends within the research area. You don’t know the general scope of the literature until you run a preliminary search. You must also sample these results for information for the background and objectives' section in the protocol.
2. Look to see if a systematic or scoping review already exists
3. Identifying preliminary MeSH/Subject Headings and keywords/keyphrases
You can identify the preliminary MeSH/subject headings, keywords, and key phrases you will need in your exhaustive search strategy. You will have more terms in your final search, but you must identify the basic subject headings and keywords you will need. This is the information that goes in your Search Methods section of your protocol. What goes in this Search Methods section of the protocol are the terms you found during your preliminary searching phase.
4. Getting test articles
5. Preparatory contacting of experts
This is the search designed by a librarian trained on how to design searches for systematic reviews. One goal of an exhaustive search is to identify all publications and as much grey literature as possible that meet study requirements. At least three databases should be utilized to conduct an exhaustive search. Another goal is to document and report the exhaustive search in such a way that it can be replicated for updates and reproduced by others after publication. See the Texas Medical Center Library's S.R. Database and Resources Libguide.
There are many things to consider when searching for and identifying the literature relevant to the research question. Researchers should be familiar with the study types and their uses for systematic reviews.
Randomized Clinical Trials (RCTs): A study where there is an initial research question, and there are controls for population and the intervention to be studied. 'Randomized' refers to the population participants, where groups are divided into categories that either receive the studied intervention and those who receive the contemporary treatment or placebo.
Dissertations and unpublished research: The importance of searching for dissertations come into play when there is little research on a new or emerging topic. Additionally, within the published research there may already be problematic bias. Searching within these resources again helps reduce bias within the review.
Non-English studies: There may also be geographical bias in studies. It is important to search non-English studies to identify all the relevant literature for a topic.
Hand searching identifies grey literature like conference proceedings, abstracts for posters, and presented papers not indexed in online databases. Sources to hand search include: subject specific professional association websites; included studies' bibliographies; and topic review bibliographies. Hand searching can be done within research centers such as archives that house information resources out of circulation. It is important to document all sources searched by hand. Information regarding the number of hand searched resources used and where resources are located should be reported within the PRISMA Flow Diagram.
Some systematic reviews involve a meta-analysis. There may be questions regarding the data produced from the study. Contacting experts to clarify data findings helps develop better accuracy and transparency of the information that will be generated from the systematic review.
There are some systematic reviews, which require image analysis. Images that are used within published journals may not be easily synthesized or examined. The quality of the original image published in the journal article could make analysis difficult. Contacting the researchers for the raw data images will help the systematic review team better conduct image synthetization and analysis.
A systematic review can take 1-2 years. After the search strategy is developed the search has to be deduplicated. When a search is reran, some new articles may be retrieved. Being in contact with the experts helps the systematic review team know what other research is out there, so that information can be put into the review for the principle of transparency.
It is important to search three to five databases. The databases and how many you choose to search depends on the topic of the systematic review. Ultimately, be sure you consult with a librarian or information specialist before undertaking your searches for studies.