Systematic Reviews
The ETH Library supports you in the preparation of systematic reviews. From formulating your research question and developing a search strategy to accessing full texts, we are at your side with expertise and resources.
What is a systematic review?
In systematic reviews, scientific literature is searched for, selected, evaluated and summarised in a transparent manner to make decisions based on the best possible evidence. The aim is to identify, evaluate and summarise all relevant empirical evidence that fulfils the predefined inclusion and exclusion criteria in order to answer a specific research question. The use of standardised and transparent methods minimises bias, producing reliable results that serve as a basis for informed decisions.
The hierarchy of evidence - often presented in the form of a pyramid - categorises the strength of evidence. The highest quality evidence is at the top, while the strength of evidence decreases towards the bottom.
A complete, transparent and reproducible literature search is the basis for systematic reviews, meta-analyses and related empirical research methods. A high-quality publication can only be produced if all relevant evidence is included in the analysis and the results.
Evidence synthesis comprises three main steps: preparation, retrieval and screening, as well as synthesis and write-up. The ETH Library offers targeted support in the key sub-steps highlighted in grey in the following diagram.
The external page PRISMA 2020 checklist can provide valuable guidance for the structured documentation of the entire process.
As with other scientific work, the starting point for every systematic review is the research question. So-called frameworks help to organise the research question into thematic blocks and formulate a search strategy. The PICO framework (Population, Intervention, Comparison, Outcome) is used particularly for clinical questions.
In addition, there are various other frameworks for different contexts, e.g:
- PEO (Population/Problem/Patient, Exposition, Outcome) for qualitative research topics.
- SPIDER (Sample, Phenomenon of interest, Design, Evaluation, Research Type) is useful for qualitative or mixed research topics that focus on ‘random samples’ rather than populations.
- SPICE (Setting, Perspective, Intervention, Comparison, Evaluation) is suitable for qualitative research topics that evaluate the outcomes of a service, project or intervention.
- ECLIPSE (Expectation, Client group, Location, Impact, Professionals, Service) for qualitative research topics that analyse the results of a policy or service.
In addition to the systematic review, there are other review types. This article provides an overview: external page A typology of reviews. An analysis of 14 review types and associated methodologies. In: Health Information & Libraries Journal, 26.2, (Jun 2009), pp. 89-168.
The search for previously published literature helps to familiarise oneself with the research topic and its scientific context. The ETH Library offers access to numerous bibliographic databases. It is also advisable to check whether other research groups are producing or have produced similar systematic reviews. This may be of interest for (international) collaborations, in order to avoid duplication of effort for existing reviews or to scrutinise the scientific relevance of the topic. Useful sources of information for searching for systematic reviews are:
- external page Cochrane Collaboration – Systematic Reviews in Healthcare
- external page Campbell Collaboration – Systematic reviews on social interventions
- external page Collaboration for Environmental Evidence – Systematic Reviews in the Environmental Sciences
- external page RefHunter – Systematische Literaturrecherche – a community-driven information portal for systematic literature searches
A protocol creates a clear, structured framework for the systematic review. It covers key aspects such as the research question, the researchers, the time frame, the inclusion and exclusion criteria of the studies, the databases to be searched and the search strategy. Registering the protocol promotes transparency, reduces the risk of overlaps and helps to avoid biased results in your research.
- external page PROSPERO – international prospektive registry for systematic reviews
- external page OSFRegistries – the Open Registries Network, a free, open platform to promote collaboration
- external page Zenodo – built and run by CERN and OpenAIRE to ensure that everyone can participate in Open Science
The external page PRISMA-P Checklist helps with the preparation of a protocol. For a Cochrane Intervention Review, please refer to the external page MECIR Guidelines for the preparation of protocols.
To ensure the transparency and reproducibility of the review, it is important to document the search string. The search string should cover all aspects of the research question and be adapted to the respective databases that are relevant to your topic. A carefully developed search strategy is crucial for the quality of the systematic review. A combination of controlled and free vocabulary should be used.
The search strategy is initially created for a specific database and must then be transferred to other databases. It should be noted that the syntax, command language and controlled vocabulary differ in the various databases.
- external page PRISMA-S-Formular: Supports the development and documentation of the search strategy
- external page SR-Accelerator-Polyglot: Helps with the translation of search strings from PubMed to other databases
The ETH Library supports you in translating your research question into a precise search strategy and in reviewing existing search strategies. It also offers access to numerous bibliographic databases and provides advice on their use.
After the search, the references from the various databases are exported to a reference management programme. The subsequent deduplication of the results ensures that only unique references are used for screening. The ETH Library offers support for the following reference management programmes: Mendeley, Zotero and EndNote.
The screening process in a systematic review takes place in two stages. The previously defined criteria provide the basis for deciding whether to include or exclude references. In the first step, titles and abstracts are checked and relevant references are selected. In the second step, the full texts are analysed. Various freely accessible and fee-based tools provide administrative support for this process.
- free of charge: external page Rayyan, external page SRDR+, external page Screenatron
- subject to a fee: external page Covidence, external page EPPI-Reviewer, external page DistillerSR
Various external page PRISMA-flowcharts are available to document the entire selection process.
Additional sources such as conference papers, reports or repositories can provide valuable information. You can find this type of additional literature by searching sources such as conference websites, organisational reports or repositories of universities and research institutions. Tools such as external page Citation Chaser or external page lens.org facilitate the search for this grey literature.
The ETH-VPN gives you access to scientific publications, even outside the ETH campus. A list of the electronic resources provided by the ETH Library can be found on the media types webpage. For unavailable articles, please use the interlibrary loan form of the ETH Library.
