Skip to Main Content

Systematic Reviews

Search Overview

Systematic reviews require thorough searching for studies, forming the basis for answering research questions and creating new knowledge synthesis.

As stated in the 'Before You Start' section, it's helpful to have a well-defined question in mind and use a recognised framework to formulate it. This can make it easier to create effective search strings that can be used or modified when searching databases or repositories.

When designing a search process, precision and sensitivity must be balanced. If you have already done a scoping search, it may need to be revised and evaluated before you settle on a final search strategy that can be used across different sources. Depending on your topic, you can use smaller strategies or "filters" (also known as "hedges") as part(s) of your search strategy.

To find published literature on your topic, you need to search multiple databases. You will also need to carry out a search for grey literature and other sources, including a "handsearch" (effectively searching highly relevant journals' table of contents or journal/publisher sites directly). This should be specified in your protocol. 

In order to be transparent and reproducible, you will need to maintain a record of your decisions and the searches that you carry out, including the dates you searched, where you searched, the search string(s) you used and the results returned.

The Search Process

1. Consideration of Databases and Tools

When creating a protocol for your research project, it is important to consider the sources you plan to search and the tools available for conducting those searches. Each database or source has its own syntax, search fields, subject and time coverage, and collection of publications. Some databases have a controlled vocabulary, such as Medical Subject Headings (MeSH) in Medline, that can be used in addition to keyword searching. At CSIRO, we have access to numerous databases that can be browsed either alphabetically or by subject. Refer to each database's help guide if you're unfamiliar with its mechanics and capabilities. These guides contain search syntax, time coverage, available fields, and other details.

2. Crafting a Search Strategy

Developing a search concept table or matrix can enhance your search strategy. This table should highlight key search concepts, potential synonyms, and ways to incorporate search headings or search syntax for precision. When developing your search strategy, you should also evaluate whether each concept is necessary or not. If a concept is not well described in the literature or indexed in controlled vocabularies, it may be best to exclude it from your search strategy.

3. Utilising Specialised Tools for Health Research

For health-related reviews, tools like the Systematic Review Accelerator's 'Polyglot' can aid in translating searches. However, it is advisable to understand how each database works before using such a tool and to also verify any translations that you choose to use.

4. Assessing Search Efficacy

While database coverage will vary, you can check the efficacy of your search by having a sample set of highly relevant items. Check whether your searches retrieve these items. If not, try to determine the reason why they are not appearing in your result.

a. Tracking Search Strategy and Results

Maintain a clear record of search strategies used, search dates and the number of results obtained. This documentation is vital for reporting on your review methodology and for updating your search if required to do so.

b. Determining Saturation Point

Be mindful of the possibility of reaching saturation point in your database search. There is no predefined number of results required to be returned, but if the search fails to yield new items, reevaluate your strategy and consider moving to the next stage of your process.

5. Supplementary Search Methods

Depending on the nature of your review, augment database searches with:

  • Handsearching: Reviewing reference lists, specific journals, or contacting authors
  • Citation Searching: Tracing forwards and backwards for relevant item citations
  • Grey Literature Searching: Exploring clinical trials databases, theses, or unpublished/non peer-reviewd research

All supplementary searching needs to be documented and incorporated into the PRISMA diagram or review methodology.

6. Managing Search Results

After finalising your search strategy and obtaining results from each database and source, you should export them to a reference management tool like EndNote. You can then deduplicate the results within the reference management tool; a specific deduplication tool; or a systematic review manager. You can then begin the screening process.

7. Periodic Search Re-evaluation

For extended reviews, periodically re-run your search to capture developments. Implement date limits or filters to minimise redundant screening. Update the PRISMA flowchart or review methodology based on new findings.