Systematic reviews are the gold standard of evidence synthesis. This guide walks you through each PRISMA step with practical tips tailored for health researchers in Nepal.
What is a Systematic Review?
A systematic review is a rigorous, structured synthesis of all available evidence on a specific research question. Unlike a narrative review, which is selective and subjective, a systematic review follows a pre-defined protocol, uses comprehensive search strategies, and applies explicit inclusion and exclusion criteria. The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines provide the internationally accepted framework for conducting and reporting systematic reviews.
Step 1 — Formulate Your Research Question Using PICO
Every systematic review begins with a structured research question. The PICO framework is the standard approach: Population (who are you studying?), Intervention (what exposure or treatment?), Comparison (compared to what?), and Outcome (what are you measuring?). For example: In adults with Type 2 diabetes in South Asia (P), does structured dietary counselling (I) compared to usual care (C) reduce HbA1c levels at 6 months (O)? A clear PICO question drives your search strategy and eligibility criteria.
Step 2 — Register Your Protocol on PROSPERO
Before searching, register your review protocol on PROSPERO (the international database of prospective systematic reviews). Registration prevents duplication, ensures transparency, and protects your work from being scooped. It is free, quick, and increasingly required by journals. Include your PICO question, search strategy, eligibility criteria, and planned outcomes in your registration.
Step 3 — Develop a Comprehensive Search Strategy
Search at least three databases: PubMed/MEDLINE, Embase, and the Cochrane Library. For public health topics, also search CINAHL, Global Health, and Google Scholar for grey literature. Use MeSH terms, free-text synonyms, Boolean operators (AND, OR, NOT), and truncation symbols (*). Ask a medical librarian to peer-review your strategy using the PRESS checklist. Document every search with the date, database, and full search string.
Step 4 — Screen Titles, Abstracts, and Full Texts
Screening must be done by at least two independent reviewers. Use Covidence, Rayyan, or Endnote for efficient title and abstract screening. After the first round, retrieve full texts for all potentially eligible studies and screen these against your criteria. Calculate inter-rater agreement (Cohen's kappa) to demonstrate consistency. Resolve disagreements through discussion or a third reviewer. Document all exclusion reasons at the full-text stage.
Step 5 — Extract Data Systematically
Create a standardised data extraction form before you begin. Pilot it on two or three studies and refine it. Extract: study design, setting, population characteristics, sample size, intervention details, comparator, outcome measures, follow-up duration, and key results. Dual extraction (two reviewers extracting independently) reduces errors. For quantitative data, extract means, standard deviations, event rates, and adjusted estimates where available.
Step 6 — Assess Risk of Bias
Risk of bias assessment evaluates the methodological quality of included studies. Use the Cochrane RoB 2 tool for randomised trials, the Newcastle-Ottawa Scale for observational studies, or the JBI critical appraisal tools for qualitative and other study designs. Do not exclude studies based on quality alone — include them but consider their quality when interpreting results.
Step 7 — Synthesise the Evidence
If studies are sufficiently homogeneous in population, intervention, and outcome, you may perform a meta-analysis to pool results quantitatively. Use RevMan 5 or R (meta package) for this. Report pooled estimates with 95% confidence intervals and assess heterogeneity using I² statistics (I²>50% suggests substantial heterogeneity). If meta-analysis is not appropriate, conduct a narrative synthesis — grouping studies by outcome and describing patterns in the evidence.
Conclusion
Systematic reviews are time-intensive but enormously valuable. They inform clinical guidelines, policy decisions, and future research priorities. Starting your first systematic review can feel overwhelming, but breaking it into these steps makes it manageable. Research Sathi offers dedicated systematic review support, including protocol development, database searching, and meta-analysis — get in touch if you need hands-on guidance.
Ms. Anita Shrestha
Systematic Review Specialist — Research Sathi
Our team brings together experienced researchers, statisticians, and academic writers dedicated to helping students and professionals across Nepal produce impactful research.