The validity, usefulness and ethics of scientific experiments that rely upon animal models are increasingly being called into question—not only by animal advocates, but by those in the scientific community. But do such experiments also fail to incorporate even the most fundamental bias reduction procedures? A recent study in PLOS Biology set out to determine if scientists conducting animal experiments report basic steps taken to reduce bias in their studies.
Researchers examined over 2,500 scientific papers published between 1992 and 2011 to see whether or not the articles’ authors reported the following four measures:
- Were animals randomly assigned to control and experimental groups? This would prevent issues such as putting healthier animals in the experimental treatment group, which could make it appear that a drug is working better than it really is.
- Were the scientists who assessed the outcome of the experiments blinded regarding which animals were treated?
- Did researchers report a sample size calculation? Having too large of a sample size results in wasting of resources and animal lives, and could signify that researchers collected data until they had a significant finding.
- Did researchers report potential conflicts of interest?
Although the study’s lead author, neurologist Malcolm MacLeod of the Centre for Clinical Brain Sciences at the University of Edinburgh, says that such measures should be standard operating procedure—“nobody in science should not be doing this stuff,” he notes—results from his study suggest that very few papers make mention of these methods. Blinded outcome assessment was only reported in 30% of publications, randomization of animals to groups in 25%, conflict of interest statements were present in 12% of studies, and sample size calculations were included in less than 1% of studies examined.
Use of bias-reducing measures has become standard in human clinical trials; however these data suggest quite the opposite is true for animal experiments. While more recently-published papers reported blinding, randomization and conflict of interest statements at higher rates than earlier publications, the rate of reporting was still quite low. “We could clearly be doing a lot better,” Macleod says.
The authors speculate that this is occurring, in part, because scientific journals are not holding researchers to higher standards. The authors note, “The most parsimonious explanation of our findings is that journal editorial policies and those charged with assessing the quality of published work, including peer reviewers, have given insufficient attention to experimental design and the risk of bias, and that this has led investigators to believe that these factors are not as important as the novelty of their findings.”
What are the implications? Poorly-designed animal studies that don’t take steps to avoid bias are misleading, waste animal lives and are an impediment to scientific advancement. These findings reinforce the necessity of having the scientific community develop and use more human-focused and humane methodologies that can spare animal suffering.
What are your thoughts on the finding that most animal research studies do not report basic steps taken to reduce bias? Send your questions and comments to email@example.com. We look forward to hearing from you.
-Dr. Pam Osenkowski, Director of Science Programs
Most animal research studies may not avoid key biases
October 13, 2015
Researchers who conduct animal studies often don’t use simple safeguards against biases that have become standard in human clinical trials—or at least they don’t report doing so in their scientific papers, making it impossible for readers to ascertain the quality of the work, an analysis of more than 2,500 journal articles shows.
For more information see: Science
For additional coverage see: Nature