Figure 1 presents the major steps in conducting a systematic review culminating in grading the strength of evidence. It presents 8 steps. (1) Review protocol: Create a review protocol that presents the proposed scope of the review. Using stakeholder input, include an analytic framework, key questions, and major outcomes. The analytic framework presents review priorities considering key populations, interventions, comparisons, and major outcomes. (2) Search and selection of relevant studies: Apply robust search strategies and inclusion/exclusion criteria to locate all available studies relevant to the review. (3) Individual Study-Level Risk of Bias Assessments: For each included study, conduct a design-specific risk of bias (RoB) assessment (low, moderate, high). When appropriate, conduct a RoB rating separately for specific outcomes in a study. (4) Group Available Evidence: Group studies, aggregating evidence for each major outcome by factors defined in the analytic framework (e.g., addressing key treatment comparisons or targeting similar patient populations). (5) Outcome-level Strength of Evidence by Study Design: For each body of evidence, score relevant domains and assess strength of evidence (SOE) separately for randomized controlled trials (RCTs) and observational studies. Use quantitative or qualitative synthesis approaches, as appropriate. a. The study limitations domain (based on RoB of individual studies) is the backbone of the required domains. b. Deficiencies in the five required domains may lower the SOE grade; presence of the relationship in the three optional domains may raise it. (6) Overall Strength of Evidence: Combine design-specific SOE assessments into one overall SOE rating for each outcome. a. Ensure that the overall SOE grade based on domain scores is consistent with a more global assessment that considers the definition of each level of grading. b. Clearly report which domains contributed most substantially to the final SOE grade. (7) Summary Tables: To promote the usefulness of findings for decisionmakers, clearly report SOE grades with other important summary data in tables. (8) Transparent Reporting: Describe and explain SOE decisions in the review’s executive summary and discussion section in a manner consistent with the relationships presented in the analytic framework. Explicitly document decision rules used for combining domains into the overall SOE grade in the review’s methods section.

Figure 1Major steps in a systematic review culminating in grading strength of evidence

Note: Adapted from © G.H. Guyatt, et al. Figure 1. Schematic view of GRADE’s process for developing recommendations. J Clin Epidemiol 64 (2001) 385. Used with permission.

From: Grading the Strength of a Body of Evidence When Assessing Health Care Interventions for the Effective Health Care Program of the Agency for Healthcare Research and Quality: An Update

Cover of Methods Guide for Effectiveness and Comparative Effectiveness Reviews
Methods Guide for Effectiveness and Comparative Effectiveness Reviews [Internet].

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.