Improving the process of translational research

BMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e7837 (Published 23 November 2012) Cite this as: BMJ 2012;345:e7837
  1. H Bart van der Worp, neurologist1,
  2. Peter A G Sandercock, professor of medical neurology and director, Edinburgh Neuroscience2
  1. 1Department of Neurology, Rudolf Magnus Institute of Neuroscience, University Medical Centre Utrecht, Heidelberglaan 100, 3584CX Utrecht, Netherlands
  2. 2Division of Clinical Neurosciences, University of Edinburgh, Western General Hospital, Edinburgh, UK
  1. h.b.vanderworp{at}umcutrecht.nl

The application of reporting standards may lead to more useful animal studies

Empirical evidence that deficiencies in the reporting of clinical studies are associated with overstatement of the efficacy of the treatment under study led to the development of the CONsolidated Standards of Reporting Trials (CONSORT) statement for the reporting of randomized controlled clinical trials.1 Accumulating evidence suggests that deficiencies in the reporting of animal studies may have a similar effect.2

In response to concerns that lack of transparency in the reporting of animal studies may have made the process of translation from bench to bedside inefficient and wasteful, the US National Institute of Neurological Disorders and Stroke (NINDS) convened a meeting of major stakeholders in translational research in June this year.3 Their aim was to develop recommendations for improving the reporting of the results of animal research, and recently they published a core set of reporting standards for animal studies.3 The group comprised academic researchers and educators, reviewers, journal editors, and representatives from funding agencies, disease advocacy communities, and the drug industry. The core standards deserve close attention.

Animal studies have been important in the development of many available medical treatment strategies. However, success in an animal model does not guarantee equal success when treatments are used in human subjects. On the contrary, most treatments that have been reported as highly promising in animal models have been disappointing when tested in clinical trials.2 Successful translation of preclinical success into clinically effective treatments depends in part on reports of animal studies providing sufficient detail of the design, methods, conduct, and results to allow other researchers to evaluate (and replicate) the findings. Deficient reporting may obscure important limitations of a study, and clinical trials based on positive but unreliable findings from flawed animal studies are likely to be unsuccessful or even lead to harm.

According to the recently published core set of reporting standards, investigators should at least report on sample size estimation, whether and how animals were randomized, whether investigators were blind to the treatment, and the handling of data. The last item includes the reporting of inclusion and exclusion criteria, of animals excluded from the analyses, and of predefined primary outcome measures. For clinical investigators and readers of clinical articles, who have been familiar with the CONSORT guidelines since 1996, these recommendations may seem superfluous. However, systematic reviews of interventions tested in animal models have consistently shown that reporting is often inadequate. For example, randomization and blinding were reported in a minority of articles, and sample size calculations, specified inclusion and exclusion criteria, and information on animals excluded from the analyses were reported in an often small minority.2 4

The development of the NINDS standards builds on important previous work.5 6 7 Comparable reporting guidelines have been published before, of which the recent ARRIVE (Animals in Research: Reporting In Vivo Experiments) guidelines have probably been endorsed most broadly—more than 160 journals have now incorporated ARRIVE in their instructions for authors (www.nc3rs.org.uk/page.asp?id=1796).8 The ARRIVE guidelines were developed using the CONSORT statement as their foundation and are therefore intuitive to clinical readers. Their 20 item checklist is somewhat more inclusive than the NINDS standards.

Of course, deficient reporting alone cannot be held responsible for the failure of results from animal studies to translate to the clinic. Several other contributing factors have been proposed. Firstly, poor design or conduct of animal studies may have resulted in spuriously positive results. Secondly, lack of power, or the use of lower doses in human subjects than the doses that were most effective in animals, may have led to negative results in clinical trials. Thirdly, some animal models may not faithfully reflect disease pathology in humans.2 9 Finally, bias towards publication of positive findings rather than neutral or negative results may lead to an overestimation of the efficacy of new treatments.10

Although the endorsement of reporting standards is a start, improving the quality of reporting will not be sufficient to reduce translational failure. We believe that a broader introduction of multicenter animal studies has a role to play too.11 Multicenter randomized clinical trials have many features that maximize transparency: prospective registration of the study, publication of the protocol before data collection (including the primary measure of outcome and the data analysis strategy), many independent sites (with a possibility of comparing results between sites), and independent data monitoring. Few animal studies can currently boast these attributes. It is instructive that an early multicenter animal study identified major research misconduct in one of the centers (a very distinguished academic center).12 There is therefore a case for more multicenter animal studies to adopt the key features of clinical trials. Not only might this increase the efficiency of translational research (larger sample size, replication, and cross validation of results between centers), but it may also allay concerns about transparency.

There is only circumstantial evidence that deficient reporting of animal studies leads to overstatement of efficacy and thereby contributes to translational failure. The proof of the pudding is in the eating. Only time will tell whether improvements in the reporting of animal studies will improve translational efficiency.


Cite this as: BMJ 2012;345:e7837


  • Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: HBvdW is supported by a grant from the Dutch Heart Foundation (2010T075); PAGS had no support from any organization for the submitted work; no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years; both authors are members of the Collaborative Approach to Meta Analysis and Review of Animal Data from Experimental Studies (CAMARADES); HBvdW is a member of the Multicentre Preclinical Animal Research Team (MultiPART); no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.