How do I trace a review's pooled estimate back to its included study data?

Trace a pooled estimate by rebuilding it from the study-level effect sizes and weights, then matching each recomputed contribution to the forest plot and the original study report. The deliverable is a tracing worksheet in which every pooled value links to a study-level source.

1. Record the effect measure and model. Note whether the review used a fixed-effect or random-effects model and which effect measure (for example a risk ratio or mean difference) the pooled value represents. The Cochrane Handbook's meta-analysis chapter treats these choices as the basis for how individual studies combine, so the model determines which weights you must reproduce.
2. Extract each study's effect size and weight from the forest plot. Copy the point estimate, confidence interval, and weight column for every included study into the worksheet. These are the inputs the pooled estimate is built from.
3. Recompute each study's contribution. Multiply or otherwise combine the study's effect size with its weight as the chosen model specifies, and record the resulting contribution beside the study row. This shows how much each study moves the pooled value.
4. Sum the contributions and compare with the reported pooled estimate. If your recomputed total matches the published pooled value, the link is confirmed. If it does not, mark the row as unresolved and check the weight column, the model, and any rounding before drawing conclusions.
5. Reconcile mismatches against the study-level source. Open the original study report and confirm the effect size and sample data used in the review. A mismatch between the forest plot and the source stays flagged until you locate the specific input that differs.
6. Check sensitivity and subgroup analyses. The Cochrane Handbook describes sensitivity analyses as a way to test how pooled results respond to analytic choices. Note which studies or choices each analysis adds or removes so the worksheet explains any shift in the pooled value.
7. Interpret the traced estimate with its confidence interval. Cochrane's interpretation chapter covers confidence intervals, exact P values, and absolute versus relative effects. Record the interval alongside the pooled value so the worksheet shows the precision of the estimate, not only its central figure.

A completed worksheet lets a reader follow any pooled number back through its weight and effect size to the study that supplied it, and it isolates the exact step where a mismatch arises.