Awareness of the existence of a reproducibility and predictability crisis in biomedical science has increased in recent years. The reproducibility crisis refers to the problem that researchers struggle to replicate or reproduce scientific studies. There has been many publications reviewing why preclinical research is irreproducible and lack of predictability, pointing this to deficiencies in reporting and statistical practices. Confounding factors, which are part of the laboratory environment and will influence both the dependent and independent variables, continue to be identified, suggesting that our knowledge of their existence is far from complete. Better statistical methodology will play a central role in improving the reproducibility of science to produce robust and reproducible research. Another area of improvement is the development of novel methods to better define and assess replication success and improve predictability. Under this light, the development and introduction of new, powerful concepts for biomedical research is essential to reduce the production of non-reproducible and non-predictive data. This has immense scientific, economic and social significance. In this context, we propose that the findings and concepts from the 3Rs field can greatly help to improve biomedical research on several levels.

Therefore, the main aim of the COST Action IMPROVE is:

To establish a network which will work to refine, harmonise and promote 3Rs concepts, data and documents, in order to improve the quality of biomedical science.