A systematic review is generally accepted to be a review of studies which involves a level of scientific integrity. This can be demonstrated by being comprehensive in the studies it includes, by being rigorous in the methods used to assess the studies, by being transparent in how studies were chosen and analysed and by being objective throughout the review process.
In the context of this review, “systematic reviews” are reviews which identify themselves as systematic reviews and are published as such. They do not necessarily live up to the reputation of being comprehensive, rigorous, transparent and fair but they are likely to be trusted by readers and decision makers as such, therefore they are the subject of this research.
An empirical paper in the context of this review is one that examines one or more systematic reviews by stating the methods and results of their observations, allowing the paper’s methodological approach to be appraised or replicated. Empirical papers tend to have been subject to formal peer review in order to be accepted for publication, however this is not always the case.
An editorial paper in the context of this review is one which highlights particular issues with systematic reviews without stating the methods for the observations such as comments and letters in academic journals. Editorials may be accepted for publication without any formal peer review process.
Some editorials present themselves in empirical format, for example a “research letter”
Descriptive or cross-sectional analysis
A descriptive or cross-sectional analysis in the context of this review examines characteristics of a series of systematic reviews (e.g., reporting or methodological quality, risk of bias, bibliometrics) but does not analyse the differences that these characteristics make to the observed summary treatment effects.
This type of review is sometimes referred to as a methodological systematic review; cross-sectional survey; or descriptive analysis . Generally the aim is to provide a complete census of relevant data and often includes comprehensive searches and duplicate screening.
A meta-epidemiological analysis in the context of this review examines the impact of certain characteristics of studies on the observed effect and provides empirical evidence for hypothesised associations. Sometimes these studies analyse the differences that certain characteristics make to the observed summary treatment effects. The unit of analysis in the meta-epidemiological analyses of interest is usually a systematic review.
This type of study may provide evidence that a particular design feature causes bias. It may adjust also for confounding.
A non-systematic review in the context of this review is a literature review which does not adhere to an empirical structure (i.e., Introduction, Methods, Results and Discussion). These may be detailed and nuanced narrative reviews, but those which have not been transparent about how the included studies under discussion were derived.