Systematic reviews

RESEARCH QUESTION – the starting point of every systematic review and scientific literature analysis. Its purpose is to precisely define what the researcher wants to find out and what data will be needed to answer the question. Its formulation determines the selection of studies, databases, and the way in which the results will be interpreted.

In clinical research, the model PICO is most commonly used. It helps structure the research question in a logical and clear way:

Depending on the type and topic of the research, other models for formulating research questions are also used.

STUDY ELIGIBILITY CRITERIA – a set of clearly defined rules used to determine whether a study should be included in the review.

They should ideally be established at the initial stage of formulating the research question, when the scope of the planned review is being defined. These criteria form the basis for study selection after studies have been identified through the literature search using the previously developed search strategy.

DATABASE SELECTION affects the completeness and reliability of a review. Therefore, in the field of medical sciences, it is recommended to use established databases such as PubMed/MEDLINE, Embase and the Cochrane Library . It is also important to include grey literature, which may contain valuable unpublished or difficult-to-access research findings.

PRINCIPLES FOR SELECTING INFORMATION SOURCES:

1. Tailor the databases to the research field.
2. Consider the type of studies and the language of publications.
3. Apply the balance principle:
   • at least 2 recommended databases (e.g., PubMed and Embase) + 1 specialized database,
   • at least 2 grey literature sources.
4. Check the availability of full-text articles (some databases have embargoes or licensing restrictions).

SEARCH STRATEGY – a set of keywords, synonyms, and Boolean operators used to identify all relevant studies.

Good practices include:

• documenting the search process (strategies, dates, number of results retrieved from each database),
• creating accounts in databases (to facilitate saving searches, accessing search history, and setting alerts),
• setting up alerts to monitor newly published studies,
• applying search limits only when justified (e.g., date, language, or study type).

BUILDING A SEARCH STRATEGY IN 7 STEPS:

1. Define the research question.
2. Identify its key concepts.
3. Find corresponding terms in controlled vocabularies MeSH / Emtree).
4. Add synonyms and alternative forms of the terms.
5. Apply Boolean operators, truncation, and wildcards.
6. Adapt the search query to the requirements and specific features of the selected database.
7. Document the search strategy.

STUDY SELECTION – a key stage of a systematic review, aimed at selecting only those publications that meet the previously established inclusion and exclusion criteria.

The study selection process should be transparent, reproducible, and consistent with the PRISMA Guidelines.

The pathway from identification to inclusion is illustrated by the PRISMA Flow Diagram.

STAGES OF THE STUDY SELECTION PROCESS:

DATA EXTRACTION – the systematic collection of information from studies included in a systematic review. Data are collected according to a predefined framework, ensuring consistency, comparability, and organization.

EXAMPLE DATA EXTRACTION PROCESS:

RISK OF BIAS AND STUDY QUALITY ASSESSMENT – a key stage of a systematic review aimed at determining the credibility and reliability of the included publications. It helps assess the extent to which the results may be affected by bias and whether they can provide a sound basis for the synthesis of scientific evidence.

SYSTEMATIC BIAS (ang. bias) – a consistent distortion of study results resulting from its design, conduct, or analysis, which may lead to incorrect conclusions.

IMPORTANCE OF THIS STAGE IN SYSTEMATIC REVIEWS:

• enables the assessment of the credibility and strength of scientific evidence,
• identifies and reduces the impact of bias on study results,
• helps prevent the inclusion of studies with low methodological quality,
• facilitates better interpretation and comparison of study results,
• increases the reliability of conclusions drawn from systematic reviews and meta-analyses.

Tools for Detecting Bias and Assessing Study Quality

ANALYSIS AND INTERPRETATION OF RESULTS – a stage of the systematic review in which the data collected from the studies are organized, compared, and synthesized to answer the research question. Depending on the type of data, this may take the form of qualitativequantitative (meta-analysis) synthesis.

THIS STAGE INCLUDES, AMONG OTHER THINGS:

• collating and comparing the results of individual studies,
• identifying similarities and differences between the results,
• assessing the consistency and heterogeneity of the data,
• conducting a qualitative (descriptive) or quantitative (meta-analysis) synthesis,
• interpreting the results in the context of study quality and risk of bias,
• formulating conclusions that address the research question.

DISSEMINATION OF RESULTS – the final stage of a systematic review, involving the presentation and sharing of the findings in a clear and standards-compliant format. Its purpose is to communicate reliable conclusions to the scientific community, clinical practitioners, and other interested audiences.

THIS STAGE INCLUDES, AMONG OTHER THINGS:

• preparing a full report or scientific article,
• presenting the methodology and results in accordance with PRISMA guidelines,
• preparing tables, graphs, and diagrams (e.g., PRISMA Flow Diagram, forest plot),
• publishing the results in a scientific journal or repository,
• presenting the findings at conferences, training sessions, or seminars,
• sharing conclusions that support clinical practice and further research.