Systematic reviews and meta-analysis are at the top of what is known as the “Evidence Pyramid”. As you move up the pyramid the amount of available evidence on a given topic decreases, but the relevancy and quality of that evidence increases. Systematic reviews and meta-analysis are considered to be the highest quality evidence on a clinical topic. However, you may not always be able to find the highest level of evidence to answer your question. You then need to consider moving down the pyramid.
Source: “EBM Pyramid.” Digital Image. Eli M. Oboler Library, 27 May 2016
The following brief definitions follow the evidence pyramid above:
Systematic Review: a comprehensive survey of a topic that takes great care to find all relevant studies of the highest level of evidence, assess each study, synthesize the findings from individual studies in an unbiased, explicit and reproducible way and present an impartial summary of the findings with due consideration of any flaws in the evidence.
Meta-Analysis: a systematic, objective way to combine data from many studies, usually from randomized controlled clinical trials, and arrive at a pooled estimate of treatment effectiveness and statistical significance.
Randomized Controlled Trial: a study in which subjects are randomly assigned to one of two groups: the experimental group receive the intervention that is being tested, and the control group receive either no treatment (placebo) or standard treatment. RCTs are considered the “gold standard” in medical research, best for answering questions about the effectiveness of different therapies or interventions.
Cohort Study: Identifies participants who currently have a certain condition or receive a treatment and are followed over time & compared with another group of people who are not affected by the condition.
Case Control Study: participants who already have a certain condition are compared with people who do not.
Case Report/Case Series: reports on the treatment of individual patients with the same condition, or on a single patient.