Article type
Abstract
Background: In the absence of clinical trials directly comparing interventions, indirect comparisons are often used. The certainty of evidence (CoE) of these comparisons can be assessed with the GRADE framework for network metanalysis (NMA), and the results of this assessment should be reported in a transparent and informative way. The GRADE for NMA evidence table describes both the absolute and relative effects of interventions and the overall CoE. However, intermediate evaluations are usually not presented, which could impair the understanding of the CoE assessment process.
Aim: To present an alternative GRADE evidence profile table for NMA with open networks.
Methods: We developed a GRADE evidence table to present the CoE assessment of an indirect comparison between dupilumab and omalizumab for allergic asthma. The domains related to direct comparison (vs. placebo), on the left side of the table, were presented together with the indirect assessment domains, the overall CoE, and the estimated effect. Figure 1 shows the GRADE evidence table proposal in this work. Structured questionnaires were applied to GRADE-trained systematic reviewers to identify the preferences related to clarity and transparency of the formal GRADE NMA table and the proposed table.
Results: The proposed GRADE evidence table presents: a) the evaluation of the direct evidence for each intervention (considering the dominant first-order loop in the NMA), except for the assessment of imprecision; b) the assessment of the indirect estimate (domains of imprecision, inconsistency – if applicable – and intransitivity); and c) the CoE and effect estimate for the indirect comparison. Eight researchers answered our survey; all of them considered that the proposed GRADE table had higher transparency and more clarity in the reporting of the CoE assessment, but similar ratings were made between both GRADE tables.
Conclusion: The proposed GRADE table provides a more transparent and clear reporting of the CoE assessment. It shows the assessment of the two direct comparisons and the assessment of the indirect comparison, covering all domains and the final CoE. This framework was developed for open networks only, but it can be adapted for other types of indirect comparison networks.
Aim: To present an alternative GRADE evidence profile table for NMA with open networks.
Methods: We developed a GRADE evidence table to present the CoE assessment of an indirect comparison between dupilumab and omalizumab for allergic asthma. The domains related to direct comparison (vs. placebo), on the left side of the table, were presented together with the indirect assessment domains, the overall CoE, and the estimated effect. Figure 1 shows the GRADE evidence table proposal in this work. Structured questionnaires were applied to GRADE-trained systematic reviewers to identify the preferences related to clarity and transparency of the formal GRADE NMA table and the proposed table.
Results: The proposed GRADE evidence table presents: a) the evaluation of the direct evidence for each intervention (considering the dominant first-order loop in the NMA), except for the assessment of imprecision; b) the assessment of the indirect estimate (domains of imprecision, inconsistency – if applicable – and intransitivity); and c) the CoE and effect estimate for the indirect comparison. Eight researchers answered our survey; all of them considered that the proposed GRADE table had higher transparency and more clarity in the reporting of the CoE assessment, but similar ratings were made between both GRADE tables.
Conclusion: The proposed GRADE table provides a more transparent and clear reporting of the CoE assessment. It shows the assessment of the two direct comparisons and the assessment of the indirect comparison, covering all domains and the final CoE. This framework was developed for open networks only, but it can be adapted for other types of indirect comparison networks.