Intended for healthcare professionals

Analysis

Is animal research sufficiently evidence based to be a cornerstone of biomedical research?

BMJ 2014; 348 doi: https://doi.org/10.1136/bmj.g3387 (Published 30 May 2014) Cite this as: BMJ 2014;348:g3387

Re: Is animal research sufficiently evidence based to be a cornerstone of biomedical research?

Pound and Bracken1 have highlighted the general poor quality of animal studies, based on emerging systematic review evidence. They have also raised concerns that some interventions with little or no therapeutic benefit may proceed to clinical trials because poorly designed animal studies have led to overestimates of their potential efficacy in humans. Our recent overview of systematic reviews,2 in which we have summarized biases reported in 31 systematic reviews (including an estimated 123 437 animals) across eight disease areas, supports the concerns raised by Pound and Bracken. We found that randomization, allocation concealment, and blinding was not done in most animal studies, and that these defects in study design can lead to overestimates of the effect sizes of reported outcomes. Moreover, our analysis independently replicates the findings from a previous systematic review of stroke outcomes reported by the CAMARADES group,3 in which studies that blinded investigators to treatment groups reduced the risk of bias.

Results obtained from poorly conducted animal studies do not provide adequate justification for clinical studies in humans. The adoption of evidence-based approaches to animal research could go some way to improving the pipeline of new effective treatments.

Jennifer A Hirst
Jeremy Howick
Jeffrey K Aronson
Carl Heneghan

References
1. Pound P, Bracken MB. Is animal research sufficiently evidence based to be a cornerstone of biomedical research? BMJ 2014; 348: g3387.
2. Hirst JA, Howick J, Aronson JK, Roberts N, Perera R, Koshiaris C, Heneghan C. The need for randomization in animal trials: an overview of systematic reviews. PloS One 2014; 9(6): e98856.
3. Crossley NA, Sena E, Goehler J, Horn J, van der Worp B, Bath PM, Macleod M, Dirnagl U. Empirical evidence of bias in the design of experimental stroke studies: a metaepidemiologic approach. Stroke 2008; 39(3): 929-34.

Competing interests: We have referred to our recent publication in this field.

17 June 2014
Jennifer A Hirst
Health Care Scientist
Jeremy Howick, Jeffrey K Aronson, Carl Heneghan
University of Oxford
Nuffield Dept of Primary Care Health Sciences, New Radcliffe House, Oxford OX2 6GG