For authors

Statistical Synthesis

Analytical choices must be specified in advance, justified clinically and statistically, and reproducible from the submitted data and code.

Required elements

• Pre-specified outcomes, time points and effect measures.
• Justification of fixed-effect or random-effects models.
• Method used to estimate between-study variance.
• Heterogeneity reporting, including I², τ² and prediction intervals where appropriate.
• Clinical and methodological assessment of heterogeneity.
• Pre-specified subgroup and sensitivity analyses.
• Justified meta-regression with adequate study numbers.
• Assessment of small-study effects and publication bias when meaningful.
• Transparent handling of zero events, multi-arm studies, multiple outcomes and missing data.
• Certainty-of-evidence assessment for key outcomes.
• Software name, version, packages and complete analytic code.