IPD may be beneficial in the following situations.

•  Many studies are unpublished or published only in the grey literature.

• There is poor reporting of studies (e.g. information presented is inadequate, selective or ambiguous).

• A high proportion of individuals has been excluded from published analyses.

•  Obtaining additional longer-term outcome data beyond that reported may provide useful insights (e.g. for mortality or child development outcomes).

• Outcome measures have been defined differently across studies.

•  Time-to-event outcome measures are required.

• Multivariate or other complex analyses are required.

•  Exploration of interactions between interventions and patient-level characteristics is important.