Short Answer
A systematic review is a reproducible literature search conducted according to predefined criteria to answer a specific clinical question. The process consists of six critical stages: structuring the question with PICO, PROSPERO registration, systematic searching, double-blind screening, quality assessment, and meta-analysis.
Reporting with the PRISMA 2020 flow diagram is a standard required by most high-impact journals. Quality assessment with the Cochrane Risk of Bias (RoB 2) or NOS, and meta-analysis processes using RevMan or the R metafor package, require expert methodological support. Serteser Danismanlik provides end-to-end methodological support for these processes, drawing on a research infrastructure that conducts PROSPERO-registered systematic reviews (Hip OA CRD420261324092, Knee OA CRD420261298163) and has produced a publication in an international peer-reviewed journal.
Systematic reviews and meta-analyses are study types that represent the highest level of evidence in the medical literature. When done correctly, they directly influence clinical decision-making, guideline updates, and research policy.
However, as the name suggests, a systematic review requires an extremely systematic process.
What Is a Systematic Review?
A systematic review is a comprehensive search of the literature carried out according to predefined criteria to answer a specific clinical question. What distinguishes it from a narrative literature review is that every step is tied to a protocol and is reproducible.
A meta-analysis, on the other hand, is the statistical combination of the included studies. Not every systematic review includes a meta-analysis, but every meta-analysis is part of a systematic review.
Step 1: Build the PICO Framework
Every systematic review begins with a clear research question. The PICO framework helps you structure this question:
- P (Population): Which patient group? (e.g., adults with knee OA)
- I (Intervention): Which intervention? (e.g., total knee arthroplasty)
- C (Comparison): What is the comparison group? (e.g., physical therapy)
- O (Outcome): What outcome is measured? (e.g., pain score, quality of life)
The clearer the PICO question, the more consistent the inclusion/exclusion criteria will be.
Step 2: Registration with PROSPERO
Before beginning your systematic review, you need to register it in the PROSPERO (International Prospective Register of Systematic Reviews) database. This registration:
- Documents the originality of your study
- Contributes to preventing publication bias
- Is now a requirement for most high-quality journals
PROSPERO registration is free, but the form is detailed. Filling out the methodology section carefully prevents deviations that would otherwise need to be explained later.
Step 3: Literature Search
PubMed, Web of Science, Scopus, Embase, and the Cochrane Library are the main databases. Depending on the scope of the study, PsycINFO, CINAHL, or field-specific databases may also be added.
The search strategy is built with a combination of MeSH terms and free-text terms. This strategy must be reported in detail in the article; the principle that any researcher should be able to repeat the same search applies.
Step 4: Screening and Selection
Title-abstract screening carried out by two independent researchers, followed by full-text assessment, is the foundation of a systematic review's quality assurance.
Tools such as Rayyan or Covidence provide great convenience during the screening process. Agreement between the two researchers is measured with Cohen's kappa and must be reported.
Step 5: Data Extraction and Quality Assessment
Data are extracted from the included studies via a standardized form. Cochrane RoB 2 for RCTs and the NOS (Newcastle-Ottawa Scale) for observational studies are the most common quality assessment tools.
Step 6: Meta-Analysis (If Appropriate)
If the studies show sufficient homogeneity, statistical combination can be performed. RevMan (free, Cochrane's tool) or the meta and metafor packages in R are used for this purpose.
The forest plot is the standard for the visual assessment of heterogeneity and the presentation of effect size. Egger's test and the funnel plot must be reported for publication bias.
PRISMA 2020 Flow Diagram
With the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 update, the flow diagram showing the study selection process has become mandatory. This diagram transparently shows how many studies were excluded at which stage.
If you are planning to write a systematic review, check the target journal's PRISMA compliance requirement in advance.
To get support during your systematic review process, request a 30-minute free consultation.
Where Do People Get Stuck Most in This Analysis?
- You have defined your inclusion/exclusion criteria, but the search returns thousands of studies, and the screening strategy is unclear.
- There is disagreement between the two researchers, and Cohen's kappa comes out low.
- You cannot decide whether a meta-analysis can or cannot be performed because heterogeneity is very high.
Related Articles
Are you looking for support during the systematic review process?
Scope and fees differ for every study; we clarify them in a free preliminary consultation and provide a written proposal. Let's discuss your study topic together and define the scope.