How can things go so wrong? What can we learn from this?
The current scientific climate stimulates publications in high impact factor journals, because this is considered important for the rankings of individual scientists as well as institutions. These incentives are nowadays also called ‘perverse incentives’, as they can lead to low quality science with adverse effects for both animals and humans. Thousands of publications of animal studies have been analysed by systematic reviews, and they all show the same picture, namely that 50-80% do not mention whether randomisation and blinding procedures have been used (Ritskes-Hoitinga & Wever, 2018). This is surprising, as randomisation and blinding are thé two starting points of our scientific practice. In case there is no randomisation and blinding, there is a higher chance of getting positive results, and thus a higher chance of getting a publication in a high impact factor journal. Moreover, many publications also do not mention a sample size calculation for identifying the proper number to be used for a good statistical analysis. The ARRIVE guidelines for good reporting of animal studies have been published in 2010 (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2010) and have been endorsed by over 1000 scientific journals in 2017, however, until now they hardly had an effect on improving the quality of reporting of animal studies (Baker, Lidster, Sottomayor, & Amor, 2014; Sena, 2017). The quality of publications on animal studies in high impact factor journals is just as deficient as in lower impact factor journals, so there is no reason to believe that publications in higher impact factor journals contain higher quality scientific results (Macleod et al., 2015).
In order to improve quality of reporting of animal studies a culture change is needed, as was concluded in a roundtable with editors and funders in Edinburgh in 2017 (Osborne et al., 2017).