1 Types of reviews
Scoping reviews, meta-analyses, systematic reviews, narrative reviews, conceptual reviews.
1.1 Narrative Reviews
1.2 Conceptual Reviews
1.3 Systematic Reviews
Systematic reviews differ from conceptual or narrative reviews in that they can afford the right to make strong statements by virtue of the employed procedures: those can and should be extremely rigorous, systematic, transparent, and reproducible.
This means that as much as possible, the decisions that are taken must be clearly documented and justified, so that the inevitable biases that the research team bring to the project can be taken into account when interpreting the results. It also means that the preparation phase is vital.
1.4 Meta-analyses
Meta-analysis is a homonym, referring simultaneously to a statistical approach and to a class of systematic reviews. The statistical approach comprises techniques to quantitatively synthesize multiple estimates of the same population parameter. Some of these techniques are relatively simple (such as when synthesizing two or more correlation coefficients), and some are very sophisticated (such as when using multi-level meta-regression).
The class of systematic reviews known as meta-analyses are those systematic reviews where the research question can be answered by quantitatively synthesizing a set of estimates, and where the heterogeneity of those estimates is sufficiently low to warrant such quantitative synthesis. Confusingly, the estimates of that heterogeneity require conducting a meta-analysis (the statistical approach). This means that in systematic reviews where the reviewers aim to quantitatively synthesize multiple estimates, but where those estimates turn out to exhibit so much heterogneiety so as to preclude a statistical meta-analysis, a statistical meta-analysis is conducted nonetheless to obtain those heterogeneity estimates. The result is a systematic review that is not a meta-analysis but that does include a statistical meta-analysis.
In such situations, the reviewers have to synthesize the estimates in another way, often resorting to qualitative integrations or to using visualisations. An insightful visualisation can be a forest plot, a visualisation typically used by statistical meta-analyses to illustrate how the synthesized estimate compares to the estimates from the separate studies – but then omitting the synthesized estimate.
1.5 Scoping Reviews
Scoping reviews or evidence maps (depending on who you ask, these can be the same or slightly different) differ from most types of systematic reviews in that they don’t answer substantive research questions (note that when I use “systematic reviews”, that also includes meta-analyses). Instead, they provide an overview of the scope of the literature: in a sense, they can tell you which research questions can be answered with systematic reviews.
Where most systematic reviews synthesize the evidence itself, often aiming to provide a more conclusive answer to the same or similar research questions as the included primary studies asked, scoping reviews synthesize the metadata about that evidence. They can tell you things like when most studies were conducted; which (sub)topics received most attention when; which study designed were used and whether that was associated to (sub)topic; how studies were distributed geographically; which sample sizes were common; whether any of those variables shows trends over time; et cetera.
Scoping reviews also produce an extensive database of literature, and are an excellent starting point for focused systematic reviews. Those also become much easier to plan, since you’ll know how many studies are available. Depending on the comprehensiveness of your scoping review’s extraction, you may even be able to skip the search and screening phases of those systematic reviews, since you already know which articles to include. Especially in combination with a decentralized approach to extraction, this means that scoping reviews can enable very efficient mapping of the literature.