ROB-MEN
ROB-MEN is a tool that helps assess how missing studies or unpublished results might bias the findings of a network meta-analysis. It looks at both missing data within and across studies, and gives a rating of low, some concerns, or high risk of bias for each treatment comparison.
At a glance
Use when
Conducting or evaluating a network meta-analysis where concerns exist about publication bias, selective reporting, or missing data that could influence treatment effect estimates.
Avoid when
When no network meta-analysis is being performed, or when there is insufficient information on study availability, unpublished data, or small-study effects to support a meaningful assessment.
Inputs
Network meta-analysis data including available pairwise comparisons, study-level results, unpublished studies information, and results from network meta-regression for small-study effects.
Outputs
Risk of bias assessment (low, some concerns, high) for each network estimate due to missing evidence, presented in a structured table format (Pairwise Comparisons and ROB-MEN Table).
How it works
ROB-MEN assesses risk of bias due to missing evidence in network meta-analyses through a two-step process. First, it evaluates within-study (e.g., unavailable results) and across-study (e.g., unpublished studies) bias for each pairwise comparison. Second, it integrates these judgments with the contribution of direct evidence, small-study effects via network meta-regression, and bias from unobserved comparisons to assign a risk of bias rating (low, some concerns, high) to each network estimate. The tool is implemented within the CINeMA framework and supported by an R Shiny web application.
- HTA domains
- Clinical Effectiveness
- Assumptions
- The availability of sufficient data to assess both within- and across-study missing evidence; the ability to model small-study effects via network meta-regression; and the validity of assumptions about the impact of unobserved comparisons.
- Strengths
- First tool specifically designed to assess risk of bias due to missing evidence in network meta-analysis.,Applicable to networks of any size and structure.,Integrated into the CINeMA framework and supported by a user-friendly R Shiny application.
- Limitations
- Relies on availability of data on unpublished studies and within-study missing outcomes.,Assessment of unobserved comparisons may involve subjective judgment.,Dependence on the quality of network meta-regression models for small-study effects.
- Also known as
- Risk Of Bias due to Missing Evidence in Network meta-analysis
Questions this answers
- › How might missing or unpublished studies affect the results of a network meta-analysis?
- › What is the risk of bias for each pairwise comparison due to missing evidence?
- › How do small-study effects influence the robustness of network estimates?
- › What is the contribution of direct evidence to the overall network results?
- › Are there unobserved comparisons that could introduce bias?
- › How can risk of bias from missing evidence be systematically reported in network meta-analyses?
References & sources
Similar by meaning
Beta record. Generated from the primary source via AI extraction and independent audit, pending final human review.