How to Write a Systematic Review

Systematic reviews represent the gold standard in synthesising research evidence across disciplines. Unlike traditional literature reviews, they employ rigorous, transparent methodologies to minimise bias and provide comprehensive answers to specific research questions. For academics and PhD students embarking on their first systematic review, understanding the structured approach is essential for producing high-quality, publishable research.

Understanding What Makes a Review 'Systematic'

A systematic review distinguishes itself through its methodical, replicable approach to identifying, evaluating, and synthesising existing research. This methodology ensures that findings are not simply the author's subjective selection of studies, but rather a comprehensive, unbiased examination of all available evidence on a particular topic.

The systematic nature encompasses pre-defined eligibility criteria, explicit search strategies, standardised data extraction, and rigorous quality assessment. These elements collectively minimise selection bias and enable other researchers to reproduce your search and verify your conclusions—hallmarks of credible academic research.

Developing Your Research Question and Protocol

Every systematic review begins with a clearly defined research question. The PICO framework (Population, Intervention, Comparison, Outcome) provides an excellent structure for clinical and health sciences research, whilst SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, Research type) suits qualitative studies better. For other disciplines, ensure your question is specific, answerable, and addresses a genuine knowledge gap.

Before commencing your search, develop a comprehensive protocol that outlines your methodology. Publishing this protocol through platforms like PROSPERO (for health-related reviews) or as a protocol paper establishes your intentions transparently and prevents outcome-switching or selective reporting. Your protocol should specify inclusion and exclusion criteria, search strategies, databases to be consulted, quality assessment methods, and planned synthesis approaches.

Conducting a Comprehensive Literature Search

The literature search forms the foundation of your systematic review and requires meticulous planning. Collaborate with subject librarians who possess expertise in advanced database searching—they prove invaluable in developing comprehensive search strategies.

Selecting Appropriate Databases

Choose databases relevant to your discipline. While PubMed, Scopus, and Web of Science serve many fields, discipline-specific databases like PsycINFO for psychology, IEEE Xplore for engineering, or ERIC for education research may be essential. Plan to search at least three major databases, supplemented by grey literature sources, conference proceedings, and thesis repositories to minimise publication bias.

Developing Your Search Strategy

Construct your search using controlled vocabulary (such as MeSH terms) combined with keywords and their synonyms. Employ Boolean operators (AND, OR, NOT) strategically to combine search concepts. Test and refine your strategy iteratively, documenting every modification. The final search string should be sensitive enough to capture all relevant studies whilst maintaining reasonable specificity to avoid overwhelming numbers of irrelevant results.

Screening and Selecting Studies

Study selection typically proceeds in two stages: title and abstract screening, followed by full-text review. At least two reviewers should independently screen all records, with disagreements resolved through discussion or third-party adjudication. This dual-reviewer approach significantly reduces the risk of erroneously excluding relevant studies.

Utilise reference management software like EndNote, Mendeley, or Zotero to organise retrieved records and remove duplicates. Specialised systematic review software such as Covidence, Rayyan, or DistillerSR can streamline the screening process and maintain transparent audit trails of decisions.

Document your selection process meticulously. The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram provides a standardised method for reporting the number of records identified, screened, excluded, and ultimately included, along with reasons for exclusion at each stage.

Quality Assessment and Data Extraction

Once you have identified your final included studies, assess their methodological quality using validated tools appropriate to your study designs. For randomised controlled trials, consider the Cochrane Risk of Bias tool; for observational studies, the Newcastle-Ottawa Scale; for qualitative research, the Critical Appraisal Skills Programme (CASP) checklist. Quality assessment informs the weight given to different studies and identifies potential sources of bias in your synthesis.

Develop a standardised data extraction form that captures essential information: study characteristics, participant demographics, methodological details, interventions or exposures, outcomes, and results. Pilot your extraction form on several studies and refine it before full implementation. Again, dual data extraction by independent reviewers enhances accuracy and reliability.

Synthesising Your Findings

The synthesis approach depends on your research question, included studies, and data availability. Meta-analysis—the statistical pooling of results—is appropriate when you have multiple quantitative studies with comparable outcomes and methodologies. Statistical heterogeneity must be assessed, and random-effects models may be necessary when substantial variability exists between studies.

When meta-analysis is inappropriate due to heterogeneity or qualitative data, narrative synthesis provides a systematic approach to describing and interpreting findings. Structure your narrative around key themes, populations, or interventions, critically examining patterns, relationships, and contradictions across studies.

Writing and Publishing Your Systematic Review

When writing your systematic review, adhere to recognised reporting guidelines. PRISMA remains the most widely accepted framework, with extensions available for specific review types such as scoping reviews, network meta-analyses, or reviews of diagnostic test accuracy.

Structure your manuscript logically: introduce the research problem and rationale, describe your methods transparently with sufficient detail for replication, present results systematically using tables and figures to enhance clarity, and discuss findings in the context of existing knowledge whilst acknowledging limitations.

Consider submitting to open access publishers like Ivory & Finch Publishers, where your rigorous synthesis can achieve maximum visibility and impact without access barriers. Open access publication ensures your systematic review reaches the widest possible audience, from fellow researchers to policymakers and practitioners who rely on synthesised evidence for decision-making.

Conclusion

Writing a systematic review demands time, methodological rigour, and attention to detail, but the resulting contribution to scholarly knowledge justifies the investment. By following established protocols, maintaining transparency throughout, and adhering to reporting standards, you can produce a high-quality systematic review that advances your field and establishes your expertise in evidence synthesis.