Research
Run a meta analysis
Effect sizes pooled properly, with heterogeneity and bias examined.
Fill it in
Population, intervention, comparator and outcome.
Pooling incomparable studies produces a confident wrong number.
A funnel plot and what it suggests.
Your prompt
Run the synthesis. [the question] Write the inclusion criteria BEFORE screening, and report the numbers at each stage. A synthesis where the criteria emerged during screening has chosen its result, and the flow diagram is what shows it did not. Extract effect sizes with their variances, and say what you did about studies reporting different measures. Examine heterogeneity properly and say what it means. Pooling studies that are not measuring the same thing produces a precise number about nothing, and a high I squared with a confident summary estimate is the classic failure. A funnel plot and what its asymmetry suggests, without overclaiming from a small number of studies. Forest plot, with weights visible.
Use Run a meta analysisOpens with everything above already filled in.
Why this works
A high heterogeneity statistic with a confident summary estimate is the classic failure: a precise number about nothing. This writes inclusion criteria before screening, reports the flow, examines heterogeneity, and checks for publication bias.