Reproducibility of science and developmental anaesthesia neurotoxicity: a tale of two cities

Editor—Reproducing scientific observations is the foundation of trusted discoveries. In clinical sciences, specific questions are repeatedly asked by a multitude of observational studies or trials, and the rigorous meta-analysis of these multicentre datasets leads us closer toward the application of evidence-based medicine in our patients. A similar approach is less frequently applied to basic and even translational research, where differences in experimental protocols used, species studied, and outcome measures reported make direct comparison of laboratory studies at best very difficult.¹ The seeming lack of interest in thoroughly replicating experiments produced by other laboratories can be explained by a multitude of factors, amongst which the difficulties to obtain funding and dedicated researchers to conduct highly time- and money-consuming "non-innovative" projects are probably the most important. It is increasingly acknowledged that many preclinical studies are statistically underpowered and cannot be reproduced.²³ The absence of confirmation of an experimental finding by other independent laboratories can greatly weaken the confidence in any particular observation. Importantly, many clinical trials are fuelled by translational experimental datasets, and these preclinical works should be reliable and of the highest quality to provide a solid base for human studies.

http://ift.tt/2urjMMG