Systematic Literature Review: Methods, Steps, and Best Practices

1. What Makes a Review "Systematic"?

A traditional (narrative) review is vulnerable to selection bias—the author might only choose papers that support their existing viewpoint. An SLR eliminates this bias by defining a strict search protocol beforethe search begins. If a paper meets the criteria, it must be included, even if it contradicts the author's hypothesis.

Because of this rigor, SLRs sit at the very top of the hierarchy of scientific evidence, particularly in medicine and social sciences.

2. The PRISMA Framework

The gold standard for reporting SLRs is the PRISMA framework (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). It provides a 27-item checklist and a 4-phase flow diagram that you must include in your publication to ensure transparency.

3. The 5 Steps of an SLR

Step 1: Protocol Development and Registration

Before searching, write a protocol detailing your exact research question (often using the PICO framework: Population, Intervention, Comparison, Outcome), your search strings, and your inclusion/exclusion criteria. It is highly recommended to register this protocol on platforms like PROSPERO to prevent duplication by other teams.

Step 2: Comprehensive Search

Execute your search strings across multiple databases (e.g., PubMed, Scopus, Web of Science). You must document exactly how many results were returned from each database. You will also need to perform "grey literature" searching (e.g., searching clinical trial registries or university repositories) to combat publication bias.

Step 3: Screening and Selection

Export all results to a reference manager and remove duplicates. The screening happens in two phases:

1. Title and Abstract Screening: Two independent reviewers read the abstracts to remove obviously irrelevant papers.
2. Full-Text Screening: The remaining papers are read in full to ensure they meet the strict inclusion criteria. Any disagreements between reviewers are resolved by a third party.

Step 4: Data Extraction and Quality Assessment

Extract the data into a standardized form. Crucially, in an SLR, you must assess the "Risk of Bias" (quality) of every included paper using validated tools like the Cochrane Risk of Bias tool for randomized trials or the Newcastle-Ottawa Scale for observational studies.

Step 5: Synthesis (and Meta-Analysis)

If the data is homogenous (similar interventions and outcomes), you can perform a Meta-Analysis—using statistics to combine the results into a single, highly powered conclusion. If the data is too diverse, you perform a narrative synthesis, systematically grouping the findings by themes or outcomes.

4. Common Pitfalls

• Scope Creep: Defining a research question that is too broad, resulting in 50,000 search results that are impossible to screen.
• Poor Search Strings: Missing key synonyms, leading to a review that accidentally excludes seminal papers.
• Single Reviewer Bias: Attempting to do an SLR alone. True SLRs require at least two independent screeners to ensure validity.

5. FAQ

6. Conclusion

Systematic Literature Reviews are labor-intensive but incredibly impactful. By adhering strictly to frameworks like PRISMA, you provide the scientific community with the highest level of synthesized evidence available.