Steps in systematic review (Part 2)

A systematic review involves a structured approach to gathering, selecting, extracting, and appraising evidence.

This ensures reliability, reproducibility, and comprehensiveness in research.

Crafting an effective search strategy

• Develop a reproducible search strategy by documenting all databases used and search parameters.
• Identify key concepts for your research question.
• Use frameworks like PICO (Population, Intervention, Comparator, Outcome) to structure your query.

Example: In children with autism spectrum disorder, does virtual learning improve academic performance compared to a physical classroom environment?

• Create a comprehensive search string incorporating all relevant terms and variations.
• Compare search results across multiple databases (e.g., Medline, ScienceDirect, PubMed) to ensure thorough coverage.

Study selection

• Begin by reading abstracts and titles to gauge relevance.
• Apply inclusion and exclusion criteria systematically.
• Involve at least two reviewers to minimise bias.
• Maintain a record of the selection process, documenting reasons for inclusion/exclusion.

Data extraction

• Utilise a standardised data extraction template.
• Extract relevant data while maintaining accuracy and avoiding unnecessary details.
• Data extraction requires methodological, clinical, and statistical knowledge.
• Consider study design elements such as randomisation, blinding, funding sources, and analytical methods.

Organising extracted data

• Use evidence tables for detailed study summaries.
• Create summary tables to facilitate synthesis and answer key research questions.
• The PICOTS framework (Population, Intervention, Comparator, Outcome, Timing, Study design) guides data element selection.
• Ensure consistency by using an iterative process for refining data extraction forms.

Evidence appraisal

• Critically evaluate the quality and relevance of selected studies.
• Assess the risk of bias and applicability to your research question.

Data synthesis and reporting

• Summarise findings into a coherent whole.
• Conduct qualitative synthesis and, if applicable, a meta-analysis.
• Use quantitative measures such as survival data, specificity, correlations, and slopes where relevant.
• Follow reporting guidelines such as PRISMA 2020 for transparent documentation.

Systematic searching and documentation

• Identify best search terms and apply relevant filters (e.g., year, language, full text).
• Use a reference manager to organise sources and remove duplicates.
• Document all steps using the PRISMA flow diagram to track search progress and study selection.

Guidelines and best practices

• Refer to systematic review handbooks like the Cochrane Handbook and the JBI Manual of Evidence Synthesis.
• Use checklists like PRISMA for reporting and PRESS for syntax verification.

By following these systematic steps—developing a search strategy, selecting studies, extracting and appraising data—you can ensure your review is comprehensive, reproducible, and impactful.

In the next session, we will explore how to write and present findings effectively. Until then, keep refining your review process and stay committed to evidence-based research!