Randomisation, allocation concealment and blind outcome assessment have been shown to reduce bias in human trials and have been used for over four decades to improve the quality of clinical trials. For animal studies, this has not been thoroughly investigated and quality standards and reporting guidelines are only just beginning to emerge.
Hirst and colleagues now conducted an overview of 31 systematic reviews of animal studies testing interventions across a range of disease areas and with various outcomes. They found that only 29% of studies reported randomization, 15% of studies reported allocation concealment, and 35% of studies reported blinded outcome assessment.
Meta-analysis showed that failure to randomise significantly increased effect sizes, thus representing a source of bias.
These results, which were published in PLoS One, demonstrate a clear need for randomisation, allocation concealment and blind outcome assessment in any animal trial, especially since human studies are often justified based on results from animals studies.
Guidelines such as The Animal Research: Reporting In Vivo Experiments (ARRIVE) guidelines, which are currently being developed, will help to increase the standard of animal trials. Until then, as suggested by Landis et al., all authors reporting results from animal studies should make an effort to include a minimum on randomisation, blinding, sample-size estimation and the handling of data to optimise the predictive value of preclinical research.
Hirst JA, Howick J, Aronson JK, Roberts N, Perera R, et al. (2014) The Need for Randomization in Animal Trials: An Overview of Systematic Reviews. PLoS ONE 9(6): e98856. doi:10.1371/journal.pone.0098856