Question:
How do I perform a systematic literature review?
Answer:
A systematic literature review (SLR) is more than just summarising articles. It is a structured method for collecting, evaluating, and synthesising research on a specific question. Unlike a traditional review, a systematic approach aims to be transparent, replicable, and comprehensive. Here’s how to approach one step by step. Check this page on systematic literature reviews for a free downloadable template and a good list of resources.
1. Define your research question clearly
Begin with a precise, answerable question. Frameworks such as PICO (Population, Intervention, Comparison, Outcome) or PEO (Population, Exposure, Outcome) help break complex questions into components. The clearer the question, the easier it is to build search strategies and inclusion criteria later.
2. Develop a protocol
A protocol is a plan for your review. It should specify:
- Your research objectives.
- Databases and sources to search.
- Inclusion and exclusion criteria (e.g. years, languages, populations).
- Screening process.
- Data extraction and synthesis methods.
Some fields require you to register a protocol (e.g. on PROSPERO). Even if not, drafting one improves clarity and prevents bias.
3. Conduct a comprehensive search
Identify the databases most relevant to your discipline — for example, Medline, PubMed, CINAHL, PsycINFO, Scopus, or Web of Science. Use a mix of keywords and Boolean operators (AND, OR, NOT) to create search strings. For transparency, record:
- Exact search terms.
- Dates of searches.
- Any filters applied.
This record is essential for replicability.
4. Screen studies systematically
Apply your inclusion and exclusion criteria in two stages:
- Title and abstract screening – Remove obviously irrelevant studies.
- Full-text screening – Examine remaining articles in detail.
Many reviewers use two independent screeners to minimise bias. Keep a record of how many articles were excluded at each stage and why, often presented in a PRISMA flow diagram. There are tools that can help with this and many other stages – check this helpful guide.
5. Extract and chart data
For each study included, record key information in a structured way. Typical fields are:
- Authors, year, country.
- Study design and sample.
- Key findings and outcomes.
- Strengths and limitations.
A data extraction table helps you compare studies side by side.
6. Assess quality
Not all studies carry the same weight. Apply a quality appraisal tool appropriate for your discipline, such as CASP (Critical Appraisal Skills Programme), AMSTAR (for reviews), or domain-specific checklists. Highlight issues like small sample size, weak methodology, or conflicts of interest.
7. Synthesise findings
Decide whether to synthesise narratively (grouping studies by theme, method, or outcome) or through meta-analysis (statistical pooling of quantitative data). Narrative synthesis is most common at student level. Whichever approach you choose, aim to identify patterns, consistencies, and gaps across the studies.
8. Report transparently
Follow recognised guidelines such as PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). Your write-up should include:
- The review question and protocol.
- Detailed search strategy.
- Inclusion/exclusion criteria.
- PRISMA diagram.
- Results of data extraction and synthesis.
- Discussion of implications, limitations, and future research.
Transparency is key: the reader must see how you reached your conclusions.
Common pitfalls to avoid
- Writing a narrative summary instead of following a systematic process.
- Failing to record search strategies.
- Ignoring grey literature (e.g. reports, theses, conference papers) where relevant.
- Overlooking study quality when drawing conclusions.
