The article screening process focuses on systematically narrowing down your pool of studies using the inclusion and exclusion criteria you developed in your protocol. Keeping careful records of inclusion and exclusion decisions (including reasons for exclusion at the full-text stage) supports transparency and reproducibility, and is typically required for PRISMA flow diagrams and reporting.
De-duplicate Records: As studies may appear in multiple databases, it’s essential to de-duplicate before screening.
Title and Abstract Screening: Start by reviewing the titles and abstracts of all retrieved studies. At this stage, remove any articles that clearly do not meet your criteria. This quick filter helps reduce the volume before a more in-depth review.
Full-Text Screening: Next, retrieve and review the full text of all articles that passed the initial screen. Use your criteria to make final decisions about eligibility. This step ensures that each study included in your review meets all required standards.
Develop a Screening Guide: It is recommended that you compile a short 4-6 question guide to help keep your screening team on track.
Dual Independent Screening: It is strongly recommended that two reviewers independently screen all articles. This improves reliability and reduces bias.
Conflict Resolution: If reviewers disagree, resolve discrepancies through discussion and consensus. When consensus cannot be reached, involve a third reviewer—ideally someone with subject expertise.
Bias refers to factors that can systematically distort a study’s results or interpretations, leading to inaccurate conclusions. In evidence synthesis, it's important to evaluate the risk of bias in each included study to better understand the strength and reliability of the overall findings. Conducting a risk of bias assessment enhances the credibility of your review, helps readers understand which findings are most trustworthy, and supports decisions about how much weight to give each study’s results in your synthesis.
Systematic Reviews: For systematic reviews, assessing risk of bias is a core component. Each included study should be appraised for possible sources of bias, such as flaws in study design, selective reporting, or lack of blinding. The Cochrane Handbook for Systematic Reviews of Interventions recommends presenting these assessments in a table or graphical format to promote transparency.
Other Review Types: For scoping reviews and other types of evidence syntheses, formal risk of bias assessment is often not required, but a critical appraisal of included studies is still encouraged. This may involve describing common limitations across studies, noting methodological weaknesses, or explaining potential sources of uncertainty in the evidence.
