A systematic review is a comprehensive and structured evaluation of existing research. It helps synthesize evidence and assess the strength of findings. Regardless of the field, conducting a systematic review provides a rigorous framework for evidence-based decision-making.
This article will discuss the essential steps of conducting a systematic review. By the end of this article, one can clearly understand how to conduct a systematic review effectively.
Understanding what a systematic review is
A systematic review is a methodical and structured method for reviewing existing literature on a particular topic. It uses predetermined criteria and protocols to ensure consistency and reduce bias. Its main goal is to synthesize all available evidence comprehensively. A systematic review ensures that all relevant studies are considered and that conclusions are based on a complete, unbiased literature review.
Systematic vs. narrative reviews
Systematic and narrative reviews differ significantly in their approaches and methodologies.
Traditional systematic reviews are objective, comprehensive, and protocol-driven. They follow a structured process to ensure that all relevant studies are included, regardless of their results.
This approach minimizes bias by adhering to explicit inclusion and exclusion criteria, utilizing a predefined search strategy, and systematically analyzing the studies. As such, the findings of a systematic review are more reliable and reproducible, which is best for decision-making.
In contrast, narrative reviews are more selective and subjective. Experts in the field mostly write them and focus on summarizing and interpreting studies based on the author's knowledge and experience. Narrative reviews do not adhere to strict protocols. This results in a biased selection of studies, leading to overlooking relevant research or favoring certain studies.
The primary difference between them is that systematic reviews emphasize minimizing bias and ensuring transparency, making them a more trustworthy tool for evidence synthesis.
Where and why they’re used
Systematic reviews are used in several fields and contexts, including academic research, clinical guidelines, policy development, and education. They synthesize a large body of research and provide valuable insights for evidence-based decision-making. In healthcare, for example, systematic reviews help guide clinical practice and inform public health policies.
Identifying a clear research question
Importance of a focused research question
The success of a systematic review relies on a well-defined research question. The research question is the foundation of the review and will shape the scope, the search strategy, and the data synthesis. When a focused research problem does not accompany systematic reviews, they include unstructured or irrelevant studies or leave out useful studies during the review phase, leading to futile conclusions.
A well-constructed research question ensures that the review stays on track, helping to narrow the search to only the most relevant studies. It clarifies the review's objectives, making it easier to determine inclusion and exclusion criteria and formulate appropriate data analysis plans.
Use of structured frameworks
Structured frameworks such as PICO, SPIDER, or PECO help to design a focused and clear research question.
- PICO (Population, Intervention, Comparison, Outcome) is commonly used in healthcare-related systematic reviews. It helps to define the population being studied, the intervention or exposure of interest, the comparison group (if applicable), and the desired outcome.
- SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, Research type) is often used for qualitative research questions and focuses on understanding the nature of the research phenomena.
- PECO (Population, Exposure, Comparison, Outcome) is useful when the research question involves understanding the effects of exposure, such as environmental factors, on a population.
Examples of effective research questions
- Good: "How does physical activity reduce depression in adults aged 30-50?"
This question is specific, as it clearly defines the population (adults aged 30-50), the intervention (physical activity), and the outcome (reduction in depression). It is also focused and measurable.
- Bad: "Does exercise help with mental health?"
This question is too broad, vague, and unspecific. It doesn't define the population, type of exercise, or the mental health outcomes to be measured, making it difficult to guide the review process effectively.
Developing and registering a protocol
Why a protocol matters
A protocol ensures that the review process is transparent and well-documented. It helps researchers stay on track and avoid bias by pre-defining the criteria for study inclusion and data analysis.
Where to register the protocol
To increase transparency and credibility, it is recommended that the systematic review protocol be registered on platforms such as PROSPERO or the Cochrane Library. These platforms provide visibility and allow for feedback from other researchers.
What to include in the protocol
When developing a protocol, it's essential to include the following key components:
- Objectives: Clearly define the aims of the systematic review. These objectives should align with the research question and provide a framework for the review process.
- Eligibility criteria: Specify the inclusion and exclusion criteria for selecting studies. This may include population characteristics, study design, language, and publication date. These criteria ensure that only relevant studies are included in the review, reducing bias and ensuring consistency.
- Search strategies: Outline the methods used to identify studies. This includes the databases to be searched, the search terms and keywords used, and any filters or limits applied. A well-documented search strategy ensures that the review is comprehensive and reproducible.
- Data analysis plans: Define how the data from the included studies will be analyzed. This includes specifying the methods for synthesizing the data and how the quality of the studies will be assessed.
Designing a comprehensive search strategy
Choosing relevant databases
Select databases most relevant to your topic for an effective systematic review search strategy. For healthcare reviews, use databases like PubMed and Cochrane Library. For social sciences or education, consider ERIC or PsycINFO. Including specialized sources like OpenGrey for grey literature can enrich your search. Choosing the right databases enhances the quality and comprehensiveness of your systematic review search strategy.
Using search operators and Boolean logic
Use Boolean operators (AND, OR, NOT) and filters to refine the search. For instance, using "AND" will narrow your results to include only studies that mention both terms, while "OR" will broaden the search to include studies that mention either term. This method increases the precision and comprehensiveness of your search.
Incorporating grey literature
Grey literature, such as conference papers, reports, and theses, is valuable as it provides insights not published in peer-reviewed journals. Including grey literature helps reduce bias and ensures all relevant evidence is considered.
Documenting the process
For reproducibility, it is crucial to document the search strategy, including search strings, dates, and databases used. This allows other researchers to replicate the search and verify the results.
Selecting studies systematically
Defining inclusion and exclusion criteria
Inclusion and exclusion criteria should be set based on factors like population, study design, language, and the year of publication. These criteria help narrow the studies to those most relevant to the research question.
Screening process steps
The screening process should involve at least two reviewers independently assessing each study to ensure consistency and reduce bias. Any discrepancies should be resolved through discussion or third-party input.
Using PRISMA flowcharts
Including a PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart in your systematic review is essential for mapping out the study selection process. The PRISMA diagram visually presents the number of studies identified, screened, included, and excluded at each stage. This transparent approach enhances the reproducibility of your review and helps readers understand the decision-making process behind study selection.
Extracting and managing data
What data to extract
Essential data to extract includes bibliographic information (author, title, year), study design details, outcomes, and results. This information will be used for synthesis and analysis later in the review.
Organizing extracted data
Tools like Excel, Covidence, or Rayyan can help organize the data and keep track of extracted information. These tools make it easier to manage large datasets and ensure consistency across reviewers.
Ensuring consistency
To maintain consistency, use a standardized data extraction form and check the extracted data for errors. This ensures that all studies are treated equally and that the data is accurate.
Assessing the quality of included studies
Why critical appraisal is essential
Critical appraisal is vital in a systematic review because it helps assess the strength and credibility of the included studies. By evaluating the quality of evidence, you can determine the reliability of the findings and identify potential biases that may affect the overall conclusions. This process ensures that the review is based on high-quality, trustworthy studies.
Choosing appropriate appraisal tools
It is important to use established appraisal tools to assess studies' quality effectively. Some widely used tools include the Cochrane Risk of Bias (RoB) tool, which evaluates potential biases in randomized controlled trials; CASP (Critical Appraisal Skills Programme) for evaluating qualitative studies; and the Newcastle-Ottawa Scale for assessing non-randomized studies. These tools provide structured criteria to ensure a consistent and rigorous evaluation.
Interpreting quality results
Once the studies are appraised, the quality scores should be integrated into the synthesis phase. Studies with higher quality scores will carry more weight in your review, while lower-quality studies may be treated with caution or excluded depending on your criteria. Incorporating quality scores into the synthesis ensures that the review's conclusions are based on the best available evidence.
Synthesizing and analyzing data
Quantitative vs. qualitative synthesis
When synthesizing data in a systematic review, it's crucial to determine whether a quantitative or qualitative approach is appropriate. For studies with numerical data, meta-analysis is typically used to combine results and provide a pooled estimate statistically. This approach works best when studies are similar in design and measures.
A narrative synthesis is more suitable for studies with non-numerical or heterogeneous data. This approach summarizes and interprets the findings without pooling data, offering insights into patterns, themes, or relationships. Choosing the proper synthesis method depends on the nature of the studies included in your review.
Approaches to meta-analysis
In meta-analysis, you can use either a fixed or random-effects model. The fixed-effect model assumes that the effect size is the same across all studies, making it appropriate when studies are similar in design and population. On the other hand, the random-effects model accounts for variability between studies and is used when studies are more diverse or heterogeneous in terms of methods or populations.
Visualizing results
Visual tools are essential for effectively presenting the results of your systematic review. Forest plots are commonly used in meta-analysis to display individual studies' effect sizes visually, confidence intervals, and overall pooled effects.
Summary tables can organize key study characteristics and findings for easier comparison. Thematic diagrams are helpful for qualitative reviews, illustrating the relationships between themes or concepts across studies. These visuals enhance clarity and help readers quickly grasp complex information.
Writing and structuring the systematic review report
Standard sections to include
A systematic review report typically consists of the following:
- Abstract: A summary of the review's objectives, methods, key findings, and conclusions. It allows readers to grasp the key elements of the review quickly.
- Introduction: Provides background information on the research topic. This section also includes the research question.
- Methods: Detailed description of how the review was conducted. This section ensures transparency and reproducibility.
- Results: Presents the review's findings, including a summary of the included studies, their characteristics, and the outcomes.
- Discussion: Interpretation of the results and their implications.
- Conclusion: This summary summarizes the main findings and their implications. Based on the evidence presented in the review, it may also suggest areas for future research or policy changes.
Follow reporting standards
Ensure your systematic review follows established reporting standards, such as PRISMA or MOOSE (Meta-analysis of Observational Studies in Epidemiology). Adhering to these guidelines ensures that your review meets high-quality standards.
Use visuals effectively
Incorporating charts and tables is essential for enhancing clarity in your systematic review. Tables organize study details and results, while charts visually represent outcomes and trends. Using these visuals helps readers quickly understand key findings and improves the overall readability of the review.
Common challenges and how to overcome them
Managing heterogeneity
Heterogeneity refers to differences between studies. Subgroup or sensitivity analysis can address this issue and provide more accurate conclusions.
Dealing with missing data
If critical data is missing, attempt to contact the authors for clarification. If this is not possible, note the limitation in the review.
Publication and selection bias
Publication bias occurs when positive results are more likely to be published. To reduce this, assess the risk of bias in the included studies and consider grey literature.
Keeping things organized
Systematic reviews require careful organization. Use project management tools and citation software to stay on top of the process.
Maintaining transparency and reproducibility
Keep thorough documentation
Track all decisions and changes made during the review process. This ensures that the review is transparent and can be replicated by others.
Version control and data sharing
Use tools like GitHub or OSF (Open Science Framework) to manage versions and share data with others.
Citing tools and protocols
Properly credit any tools, frameworks, or software used during the review process.
Getting expert support for your review
Systematic reviews require expertise and attention to detail. If you feel overwhelmed or want to ensure your review is conducted with the highest level of precision, consider contacting a systematic review writing service such as prosystematicreviewwritingservices.com. These experts will guide you through each step and deliver a systematic review that meets rigorous standards and is ready for publication.
