Meta-analysis

A meta-analysis takes the numbers from several studies on the same question and combines them, so the answer rests on more patients than any one study had. It is usually the statistical step inside a systematic review: the review finds and screens the studies, the meta-analysis pools their results.

Pooling is only as good as what goes in. A meta-analysis of randomized trials carries more weight than one built on cohort studies or lab work, and a pooled estimate with a wide confidence interval can still leave the question open. The strongest form is a meta-analysis of individual patient data, which reanalyzes every patient's record rather than each study's summary numbers.

Ryan's research articles list each meta-analysis in their "evidence at a glance" table with its design, how many studies it pooled and the weight it deserves, and say what it could not settle.