Glossary · Methodology

Systematic review

In short

A systematic review searches for all studies on a question according to a plan set in advance, appraises their quality and summarizes them. The plan prevents only the convenient studies from being selected.

The difference from an ordinary review article lies in the method: search terms, databases and inclusion and exclusion criteria are fixed before the results are known. Anyone who wants to check can repeat the search.

If the data can be pooled, the review contains a meta-analysis; if the studies are too different, it remains an orderly description. The best-known systematic reviews are produced by the Cochrane Collaboration; reporting standards are described in the PRISMA guideline.

A review is also only as current as its search date. Large new studies can still shift the picture afterwards.

First find all the studies, then judge — not the other way round.

Where this comes up in the fact checks

Coenzyme Q10 – The Cochrane review on heart failure rates mortality and hospital admissions as moderate-quality evidence and notes that the mortality result comes from a single trial.
Cutting down on alcohol – The analysis of 107 cohort studies shows how much the choice of comparison group matters.

Sources

All terms in the glossary

Information only, not medical advice. The BK-Score is a subjective assessment according to disclosed rules – not a scientific evaluation. Scoring rules