OBJECTIVES: A large part of medical research is observational. In order to reinforce widespread adoption of evidence from observational studies (OSs), a consensus regarding accepted principles for evaluation of their credibility and reliability was developed over the years. The aim of this work was to investigate publication trends between (i) RCTs and OSs and (ii) different study designs in published observational research over the last 38 years (1980-2018).

METHODS: Advanced research on publications in PubMed-indexed journals between 1980 and 2018 has been performed searching for “Observational study”, “Real world data”, “Real world evidence”, “Registry”, “Big Data” (for OSs) and for “Randomized Clinical Trials” (for RCTs). Studies involving Humans were filtered. As indicator of publication trends, the ratio between total number of articles related to RCTs and to OSs (obtained as described previously) was calculated.

RESULTS: Overall 241677 OSs were published from 1980 to 2018 (265 in 1980 and 18193 in 2018), with an increasing trend and a peak during 2010-2015 (n=92652).

While OSs publication trend kept growing with a 2-fold increase over the years, RCTs trend changed from a 2-fold increase (1980-1997) to a 1.5-fold increase starting from 1998. The ratio of RCTs/OSs decreased in a continuous manner from 1995 (ratio=7.4) till 2018 (ratio=1.3).

Cohort studies made up 75% of all publications in the early eighties (1980-1985) and decreased to 54% in 2010-2015, while cross-sectional studies increased from 4% (1980-1985) to 12% (2010-2015). Case-control studies represented 21% of all publications in 1980-1985 and slightly increased to 33% in 2010-2015.

CONCLUSIONS: An increasing trend of published OSs covering real world data was observed since 1980 as compared to the increase of RCTs. This is probably due to the enhancement of quality of methodologies and the increased interest in observational research, which complement and build on the evidence base established by RCTs.