Table 4 Approach for assessing GRADE domains
GRADE Domain | Means of Assessing | Example |
|---|---|---|
Risk of bias | Done in accordance with assessments from JBI Prevalence Critical Appraisal Tool | A study rated ‘high risk of bias’ will be rated as either low certainty or very low certainty in this domain, depending on extent of additional comments from reviewers |
Imprecision | Considering both the sample size and by considering the validity and rigor of the methods in the study (a question from the JBI Prevalence Critical Appraisal Tool) | A study with a sample size of less than 50 individuals that also uses an outdated assay (such as immunofluorescence assay) will be rated as very low certainty in this domain |
Inconsistency | Comparing the prevalence result of a given study with how closely it aligns with other prevalence results from the same stratum | If a study’s prevalence result diverges greatly from many others in the same stratum, that study will be rated as low certainty in this domain |
Indirectness | Based on how well the study population is described (a question from the JBI Prevalence Critical Appraisal Tool) and whether the study’s main objective was to measure prevalence of a filovirus in humans | If a study provides detailed descriptions of the study population and a prevalence survey was the main intent of the study, that study will be rated as high certainty in this domain |
Publication bias | By looking at JBI Prevalence Critical Appraisal Tool overall rating for a given study and other studies in the same stratum | If a study has itself been rated ‘high risk of bias’ and many other studies in its stratum were also rated as having concerns about bias, then the study would be rated very low certainty in this domain |