ROB-MEN

ROB-MEN assesses the risk of bias due to missing evidence in network meta-analyses by evaluating selective outcome reporting, publication bias, and their impact on NMA estimates.

Key Features:

• Assessment of missing evidence: Systematically evaluates risk of bias due to missing evidence for each pairwise comparison, including within-study missing results and across-study (unpublished study) concerns.
• Integration with NMA estimates: Integrates bias assessments with the contribution of direct comparisons to overall NMA estimates, considers possible small-study effects using network meta-regression, and evaluates bias from unobserved comparisons.
• Risk level assignment: Assigns a categorical risk level to each NMA estimate using "low risk," "some concerns," or "high risk" labels.
• Output tables: Produces analytical outputs including the Pairwise Comparisons Table and the ROB-MEN Table to summarize assessments.
• Applicability to network structures: Applies to networks of any size and geometry.
• Integration with reporting frameworks: Is incorporated into the CINeMA framework and software under the reporting bias domain.

Scientific Applications:

• Coronary artery disease diagnostics: Applied to a network of non-diagnostic modalities for detecting coronary artery disease in patients with low-risk acute coronary syndrome.
• Major depressive disorder treatments: Applied to a network of 18 drugs from head-to-head studies for the acute treatment of major depressive disorder in adults.

Methodology:

Systematically evaluate missing evidence for each pairwise comparison (within-study and across-study), integrate the contribution of direct comparisons to NMA estimates, evaluate small-study effects using network meta-regression, assess bias from unobserved comparisons, and assign categorical risk levels ("low risk," "some concerns," "high risk").