The Reproducibility Crisis: Old Habits Die Hard
It has been a while since we’ve reported on the reproducibility crisis in preclinical research, but it’s not because the problem has been getting better.
It’s been estimated that between 51% and 89% of animal studies are not reproducible. In large part this is because important details about the methodologies researchers use in their experiments are omitted from their scientific publications. A recent article in Science discussed this issue—and identified a potential solution for this problem of sloppy reporting on animal experiments.
In 2010, the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs) tackled this issue head on and developed the ARRIVE (Animal Research: Reporting of In Vivo Experiments) checklist, a specific set of 38 guidelines to help improve the standard of reporting on research using animals.
Although these guidelines were endorsed by more than 1,000 scientific journals and dozens of funding agencies, studies have shown that there has been almost no improvement in how the animal experiments are being reported—even in journals that endorsed the guidelines.
Now, a streamlined version of this checklist, dubbed ARRIVE 2.0, has been created and published in seven scientific journals to improve the quality of reporting. The new checklist focuses on the “Essential 10” details that animal studies need to report, as well as 11 recommended items of secondary importance.
The guidelines include recommendations on how to report on study design, experimental procedures, animal usage and more. Some “essential” items that must be reported include the number of animals used in an experiment, their sex, whether the animals were randomly assigned to control and experimental groups, and whether the researchers in the study were blinded as to which groups the animals were in. The guidelines were created to ensure that research findings are communicated to the scientific community in an accurate and transparent way.
However, concerns remain about enforcing these guidelines, according to some journal editors who, in the past, have received incomplete checklists or ones with information that doesn’t match what is in the corresponding manuscript. Other editors note that researchers may not have an incentive to follow the guidelines—although that may change if journals refuse to publish papers that don’t comply.
Of course, these types of articles—and the problems addressed therein—bring to light a more pressing issue: why the research community should have the luxury of continuing to work with animal models at all.
Issues with animal experiments extend beyond study design and transparency and include inherent problems such as extrapolating data across species. If the scientific community really wants to improve human health and well-being through research, then less effort needs to be expended on “fixing” what’s wrong with animal models, and more effort needs to be made on developing and using human-relevant models that have a higher probability of improving human health.
Help NAVS continue to support smarter science that advances discovery, innovation and human-relevant solutions without the use of harmful, flawed and costly animal experiments by making a donation today.