Term: Systematic Review and Meta-Analysis
A systematic review is a type of research that “summarizes the results of available carefully designed health care studies (controlled trials) and provides a high level of evidence on the effectiveness of health care interventions.”
The key to this type of research is the methodologic rigor. A review article without this rigor is subject to bias, including publication bias and the bias of the author(s). For example, we all know that there is a statistic to support almost any view in medicine. If one were to pick and choose articles that support a preconceived view, a review could lead to erroneous conclusion. A systematic review is designed to include ALL research on a specific question and thus help limit this type of bias.
If there are enough studies on a particular topic with a similar enough design, some researchers may also choose to do a meta-analysis. This involves weighting studies and combining the statistics into a single value. This can help readers better interpret the meaning of multiple studies. All meta-analyses should start with a systematic review, but not all systematic reviews necessarily should lead to meta-analysis.
The major caveat to interpreting systematic reviews and meta-analyses is to assess this rigor. “Garbage in --> Garbage out” is a common mantra to describe the high number of inappropriate meta-analyses that are done in medical literature. Readers of these studies need to be careful to assess for bias in search strategies and original studies. A well-done systematic review will have a risk of bias analysis and assessment of heterogeneity of studies included.
Another important question for meta-analyses is whether the studies should have been included in a single analysis in the first place. Are they on similar populations? Are they the same study type? Are the interventions similar? Does one study grossly overpower all of the others?
Several checklists and tools exist to help readers assess quality of systematic reviews and meta-analyses. These include PRISMA (Preferred Reporting in Systematic Reviews and Meta-Analysis) and MOOSE (Meta-Analysis of Observational Studies in Epidemiology), and the Institute of Medicine Checklist. The Cochrane Collaboration is often regarded as the expert in this topic. They publish many systematic reviews on a variety of medical topics and also have resources on their website for researchers as well.