The hierarchy of scientific evidence is a concept from the field of evidence-based medicine. Most physicians strive to offer care that is grounded in scientific evidence, but while it is generally accepted that randomized controlled trials (RCTs) are the gold standard for support of a medicinal claim, the differences in scientific strength between other types of studies is less obvious. Beginning in the 1980s, various committees around the world have generated rankings of types of studies that serve as basis for professional guidelines and regulations. Rankings are not uniform, but the differences are very small.
The most often-used ranking system for scientific evidence is that of the Oxford Centre for Evidence-based Medicine
a variation on which is offered by the American medical association.
A simplified version is provided by the Australian government department of health in this diagram.
RCTs are an established standard for evidence-based medicine. Participants in such studies are recruited to achieve a sample that is homogeneous with respect to pre-defined characteristics, and are allocated in a random manner to receive either active therapy, control therapy or placebo. Such studies may employ other design aspects aimed at reducing bias, including masking the treatment assignment from the researcher, from the participant or both.
An even higher level of evidence is a number of RCTs looking at a similar medical question, with similar designs and similar populations. A systematic review of such a group of studies can provide an overview of whether a certain result is strong enough or whether it was unique to a specific study and was not replicated in other studies. The Cochrane reviews are a very good example of a collection of such scientifically-vigorous systematic reviews.
Thus, the topmost level of scientific evidence is the systematic review, a subtype of which is the meta-analysis that can provide a quantitative result of the combination of analyzed studies.
Lower in the hierarchy are cohort studies, which tend to form a sizeable proportion of the types of research conducted by younger medical professionals. These are studies in which pre-defined populations are followed over time with the attempt of finding certain parameters that are correlated to a certain outcome. For the most part, such studies are conducted retrospectively, by analyzing medical records and databases containing data that has been collected as part of the patient’s routine care. Provided access to the right database, such studies can have the advantage of a very large amount of data spanning many years of follow-up. However, cohort studies can also be conducted in a prospective manner, with the distinct advantage that the data of interest is collected at pre-defined intervals in a uniform manner. Well-controlled and designed case-series, especially those with a relatively large sample, in which each case (patient) has one or more adequate controls, are on a similar level of scientific evidence as cohort studies.
Last, but not least in the hierarchy of levels of scientific evidence are medicinal claims based on the clinical experience and opinions of experts, or on the extrapolation of non-clinical data (in-vitro studies and/or animal research).
In your writing, you will need to be aware of these distinctions when writing the Introduction to your papers, to understand which claims have been well-enough supported to form a strong basis to base your own work on. You would also need to take these definitions into account when writing your Discussion section, when comparing your results to others’ in your field.