How do I fix a pooled analysis that omitted a study from the forest plot?

Compare the included-study list with the forest plot rows to confirm the omission, add the missing study, recompute the pooled estimate, and check whether the conclusion changes.

1. Reconcile the two lists. Match every study in the included-study table against a row in the forest plot. Record each study that appears in the table but has no row, and note its effect estimate and variance or standard error. The expected result is a short list of omitted studies with the data needed to add them.
2. Add the missing study to the analysis. Enter its effect estimate and precision into the same effect measure and model (fixed or random effects) used for the other studies. The expected result is a forest plot whose rows match the included-study list.
3. Recompute the pooled estimate and heterogeneity. Report the new pooled effect, its confidence interval, and the heterogeneity statistics alongside the original values. The expected result is a clear before-and-after comparison.
4. Run a sensitivity analysis. Recompute the pooled estimate with and without the added study, and check whether the direction, size, and confidence interval of the pooled effect change. The expected result is a statement of whether the added study shifts the result materially.
5. Check the conclusion against the new estimate. Compare the recomputed confidence interval and exact P value with the review's stated conclusion. If the added study changes the direction or the interpretation of the effect, revise the conclusion; if it does not, state that the conclusion is unchanged.
6. Document the correction. Record the omitted study, the reason it was missed, the original and corrected estimates, and the effect on the conclusion in the review record or an erratum. The expected result is an auditable trail that lets readers see what changed and why.

The Cochrane Handbook's meta-analysis chapter covers effect measures, fixed and random effects, heterogeneity, and forest plots, and its interpretation chapter covers confidence intervals, exact P values, and responsible review conclusions. Use those chapters to keep the recomputation consistent with the methods already reported.