The Extent of the Use of GRADE in Campbell Systematic Reviews: A Systematic Survey

Objective To conduct a systematic survey on the extent of the use of the grading of recommendations, assessment, development, and evaluations (GRADE) and other evidence grading systems in Campbell systematic reviews (SRs). Study Design and Settings Campbell SRs published before January 1st, 2024, that used evidence grading systems were included. General characteristics and details of a summary of findings (SoF) table and an evidence profile (EP) were independently extracted by two investigators. Results Among 234 SRs retrieved, 46 (19.66%) used evidence grading systems, all of which were GRADE. One used GRADE erroneously to rate the quality of individual studies rather than the body of evidence. The 45 SRs used GRADE to assess the certainty of a body of evidence and included 858 outcomes. Of them, the certainty in evidence for 32 were rated as high (3.73%), 170 were moderate (19.81%), 291 were low (33.92%), and 365 were very low (42.54%). Among the 1860 instances of downgrading and upgrading, the certainty in evidence was mostly downgraded for risk of bias (ROB) (1026, 55.16%) and imprecision (408, 21.94%). The large magnitude of effect (14, 0.75%) and plausible confounding (10, 0.54%) were the main upgraded factors. The proportions for higher certainty in evidence (including high and moderate) were larger in the international development (9.59%) and social welfare (7.55%) groups than in the other groups (1.37%). Conclusion Most Campbell SRs do not assess the GRADE certainty in evidence. Where evidence is evaluated, the quality of that evidence is mainly low or very low, and this is most commonly due to serious ROB or imprecision.