Meta-Analyses [A]

Meta-Analyses: How We Read Evidence HereЗакреплено в сообществе

This community publishes only research syntheses: systematic reviews and meta-analyses of psychoactive substances, medicines, and drugs used in sport. It does not publish isolated studies or anecdotes as editorial posts. Each thread begins with a clearly defined question: one substance, medicine, or substance class; a defined population; and a specific outcome. That does not mean every thread covers only one molecule; it means the question is specific enough to be tested.

A systematic review defines its question and eligibility criteria in advance, searches for relevant studies, appraises bias, and synthesizes evidence using explicit, reproducible methods. A meta-analysis is the statistical pooling of study estimates when pooling is appropriate. A systematic review may contain no meta-analysis, so a review without a pooled number is still a review.

For included randomized or observational studies, we report blinding when documented, sample size, follow-up, effect size, confidence intervals, inconsistency, and risk of bias. We distinguish association from causal effect. Confidence in a finding depends on risk of bias, consistency, directness, precision, and publication bias. A pooled estimate does not automatically mean high-certainty evidence. Meta-analysis often carries more evidential weight than a single study, but poor inputs can still mislead.

When a later review adds meaningful evidence, we add a dated update to the relevant thread and state whether the interpretation changed. No dosing, procurement, use instructions, or personal medical advice. Bring a review, a methods question, or a source-backed correction.

https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-01

https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-10

https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-14

↑ 1

Обсуждение

Комментарии

0

Комментариев пока нет.

Написать Join — чтобы участвовать в обсуждении.