SYSTEMATIC REVIEW –FREQUENTLY ASKED QUESTIONS –Part 1 & 2

Systematic reviews and literature reviews differ in purpose, structure, search rigor, and how bias is handled. A systematic review is built to answer a specific research question using a protocol-driven, explicit, and accountable methodology. That structure shows up in comprehensive, reproducible search strategies; mandatory quality assessment of included studies; and bias-reducing steps such as double data extraction and systematic appraisal. Literature reviews, by contrast, are typically more flexible and narrative, often using more selective and variable searching, with less emphasis on formal bias control and quality assessment—meaning they are more vulnerable to bias. Systematic reviews also use defined data synthesis methods—ranging from meta-analysis to qualitative approaches like metaethnography and mixed-method synthesis—whereas literature reviews often rely on narrative or qualitative summaries, even when meta-analysis appears in some cases.

The transcript then tackles practical “how many” questions that frequently stall review teams. There is no single correct number of databases or total sources to search; choices depend on the research question, discipline, and how broad the scope is. Clinical trial reviews commonly draw on Medline, Embase, and the Cochrane Central Register of Controlled Trials, while nursing and allied health work may add CINAHL. Education research often uses ERIC, psychology may use PsycINFO, and sociology may use Social Science Abstracts. Beyond databases, teams are urged to search trial registries, organizational websites, and to use hand-searching and reference list scanning. Regional databases matter too—for example, when studies come from India or Latin America.

Similarly, there is no fixed number of studies that must be included. A systematic review can include as few as two or three studies for a narrow question, while broad evidence gap maps can include hundreds. Even “zero studies” can be a legitimate and valuable outcome because it signals a research gap and guides future primary research. Across all scenarios, the emphasis stays on quality over quantity: eligibility criteria should be transparent and grounded in frameworks such as PICO.

Team composition is also flexible but not optional. At least two independent reviewers are needed for screening and data extraction to reduce bias, with an adjudicator recommended to resolve disagreements. Information specialists (librarians) help design comprehensive search strategies across many databases, and methodological/statistical expertise supports meta-analysis and other synthesis decisions. Subject-matter expertise helps ensure relevance and accurate interpretation.

Protocol registration is strongly recommended to improve transparency and credibility, reduce duplication, and provide a “road map” for the team. The transcript lists common registries such as PROSPERO (health-focused) and OSF, along with options through Cochrane, Campbell, and other platforms.

Finally, the transcript distinguishes systematic maps (evidence gap maps) from systematic reviews. Mapping is suited to broad, multifaceted questions and does not require detailed extraction of study findings or formal synthesis; it often uses categorization and clustering to identify where deeper systematic reviews are warranted. It also introduces tools and methods for later stages of review work—software for managing systematic reviews, approaches to synthesis (aggregation vs configuration), embedded methodological studies (SAR), quality appraisal tools for both primary studies and reviews (e.g., AMSTAR 2, ROBIS, CASP), reporting standards (PRISMA and extensions), and priority screening using machine learning to speed title/abstract screening. A second segment continues with these methodological and operational details, including when mixed-method systematic reviews are appropriate and how to assess risk of bias and reporting quality.