How to Conduct a Nursing Systematic Review (Step by Step)

What Is a Systematic Review?

A systematic review is a rigorous, transparent, and reproducible method for synthesizing research evidence on a specific question. Unlike narrative literature reviews, systematic reviews follow predefined protocols, explicit inclusion criteria, and standardized methods to minimize bias.

Systematic reviews sit at the top of the evidence hierarchy and are foundational to evidence-based nursing practice. They inform clinical guidelines, policy decisions, and future research directions.

Step 1: Define Your Research Question

A well-defined question is essential. Use frameworks to structure your question:

PICO(S) Framework

• P - Population: Who is being studied?
• I - Intervention/Exposure: What is being evaluated?
• C - Comparator: What is it compared to?
• O - Outcome: What outcomes are measured?
• S - Study design (optional): What types of studies will you include?

Questions Suited for Systematic Review

• Effectiveness of interventions
• Accuracy of diagnostic tests
• Prognosis or risk factors
• Patient experiences (qualitative systematic review)
• Prevalence or incidence

Step 2: Develop and Register Your Protocol

A detailed protocol ensures transparency and reduces the risk of bias from post-hoc decisions.

Protocol Components

• Research question and objectives
• Inclusion and exclusion criteria
• Search strategy (databases, terms, limits)
• Study selection process
• Data extraction procedures
• Quality appraisal tools and process
• Data synthesis methods

Protocol Registration

Register your protocol with PROSPERO (International Prospective Register of Systematic Reviews) before beginning your search:

• Increases transparency and credibility
• Prevents duplication of effort
• Reduces risk of publication bias
• Required by many journals for publication
• Free registration at crd.york.ac.uk/PROSPERO

Step 3: Develop Your Search Strategy

Select Databases

For nursing systematic reviews, search at minimum:

• CINAHL: Primary nursing database
• PubMed/MEDLINE: Comprehensive biomedical literature
• Cochrane Library: Existing reviews and trials
• Embase: European biomedical literature (if accessible)
• PsycINFO: For mental health topics
• Dissertation databases: ProQuest, NDLTD

Build Search Terms

• Use controlled vocabulary (MeSH terms, CINAHL headings)
• Include synonyms and related terms
• Use Boolean operators (AND, OR, NOT)
• Apply truncation for word variations
• Test and refine your search iteratively

Sample Search Strategy Structure

1. Population terms (nurse* OR "registered nurse" OR RN...)
2. Intervention terms (intervention terms OR synonyms...)
3. Outcome terms (outcome terms OR synonyms...)
4. Combine: #1 AND #2 AND #3
5. Apply limits (English, date range, study type if appropriate)

Supplementary Searching

• Hand-search key journals
• Check reference lists of included studies
• Search gray literature (conference proceedings, reports)
• Contact experts in the field
• Search clinical trial registries

Step 4: Select Studies

Screening Process

1. Remove duplicates: Use reference management software
2. Title/abstract screening: Two reviewers independently screen all records
3. Full-text review: Obtain and assess potentially relevant studies
4. Resolve disagreements: Discussion or third reviewer

Inclusion/Exclusion Criteria

Define clear, specific criteria before screening:

• Population: Age, setting, condition, etc.
• Intervention: What counts as the intervention?
• Comparators: What comparisons are acceptable?
• Outcomes: Which outcomes must be reported?
• Study design: Which designs will you include?
• Language: English only or others?
• Date range: Publication limits?

PRISMA Flow Diagram

Document your selection process using the PRISMA flow diagram showing:

• Records identified from each source
• Duplicates removed
• Records screened and excluded
• Full-text articles assessed and excluded (with reasons)
• Studies included in final synthesis

Step 5: Assess Study Quality

Not all studies contribute equally. Quality appraisal evaluates risk of bias.

Quality Appraisal Tools

• RCTs: Cochrane Risk of Bias tool (RoB 2)
• Non-randomized studies: ROBINS-I
• Observational studies: Newcastle-Ottawa Scale
• Qualitative studies: CASP, JBI checklist
• Mixed studies: MMAT (Mixed Methods Appraisal Tool)

Quality Appraisal Process

• Two reviewers independently assess each study
• Use standardized tool appropriate for study design
• Resolve disagreements through discussion
• Document quality ratings for all studies
• Consider how quality affects synthesis and conclusions

Step 6: Extract Data

Data Extraction Form

Develop a standardized form including:

• Study identification (authors, year, title, journal)
• Study characteristics (design, setting, sample size)
• Participant characteristics (demographics, inclusion criteria)
• Intervention details (components, duration, intensity)
• Comparator details
• Outcomes (measures, timing, results)
• Quality appraisal ratings

Extraction Process

• Two reviewers independently extract data
• Compare extractions and resolve discrepancies
• Contact study authors if data is unclear or missing
• Pilot test form on first few studies

Step 7: Synthesize Findings

Narrative Synthesis

When studies are too heterogeneous for meta-analysis:

• Describe studies systematically
• Organize by theme, population, or outcome
• Compare and contrast findings
• Identify patterns and outliers
• Use tables to summarize key study characteristics

Meta-Analysis

Statistical combination of results when appropriate:

• Studies must be sufficiently similar (clinically and methodologically)
• Calculate effect sizes (OR, RR, SMD depending on outcome type)
• Test for heterogeneity (I² statistic)
• Choose fixed or random effects model
• Create forest plots
• Conduct sensitivity analyses
• Assess publication bias (funnel plots)

GRADE Assessment

Assess certainty of evidence using GRADE (Grading of Recommendations Assessment, Development and Evaluation):

• High: Very confident in effect estimate
• Moderate: Moderately confident
• Low: Limited confidence
• Very low: Very little confidence

Step 8: Report Your Review

PRISMA Reporting Guidelines

Follow PRISMA 2020 statement for transparent reporting:

• Title indicating systematic review
• Structured abstract
• Registration information
• Complete search strategy for at least one database
• Flow diagram of study selection
• Characteristics of included studies
• Risk of bias assessment
• Results for all outcomes
• Discussion of limitations
• Implications for practice and research

Common Sections

• Introduction: Rationale, objectives, research question
• Methods: Protocol, search strategy, selection, appraisal, synthesis
• Results: Search results, study characteristics, quality, synthesis
• Discussion: Summary, comparison to literature, limitations, implications
• Conclusion: Key findings and recommendations

Timeline and Resources

Systematic reviews are resource-intensive:

• Timeline: 6-18 months typically
• Team: Minimum 2-3 reviewers
• Librarian: Essential for search strategy development
• Software: Covidence, Rayyan, RevMan, or similar
• Statistician: If conducting meta-analysis