A good father loves all of his children equally. Does it stand to reason that an editor should like equally all kinds of research, or at least all that fall within his or her journal’s remit?
If so, I may be falling short of an aspirational standard. Having now spent the better part of a decade as the Editor-in-Chief of a large, international journal, and having seen more than 12,000 manuscripts go by, I will confess to being uneasy about animal research.
Some preclinical research papers describe animal models that seem needlessly painful or inhumane; frankly, I wonder how some received ethical approval. In my experience, though, this isn’t the largest part of the problem. My larger concern is that I have trouble imagining a path from bench to bedside for most animal models in orthopaedic surgery. That being so, I find myself wondering why animals’ lives are used in these ways, and whether they should be.
In this month’s Clinical Orthopaedics and Related Research®, Alper Öztürk MD and Önder Ersan MD, two clinician-scientists from Ankara, Turkey, put some numbers to this concern . They found that more than 40% of papers based on animal models that were presented at the national orthopaedic congress of their country (population 83 million) over a 9-year span were never published, and of those that were, nearly 40% were never cited or were cited only once. All of this nonimpact cost more than 9400 animals their lives. Although I have not seen this kind of analysis done before in our specialty, complementary analyses have been done in related disciplines, with similar results. For example, one study that analyzed papers published in the two clinical journals with the highest Impact Factor in each of 10 surgical specialties found the median number of citations of animal research papers by subsequent human/clinical research over a 10-year span was only one (with the high end of the range being five), suggesting minimal translation of animal studies to research in humans .
While I think citation analysis is a crummy surrogate for actual research quality, it’s a fair way to assess whether other scientists believe a piece of research is meaningful. And to put things in context, one citation per article over 10 years is a deplorably low number; a journal’s Impact Factor is calculated (more or less) as the number of citations per article over a 2-year period. CORR’s Impact Factor last year was just over 4, meaning slightly more than four citations per paper over a 2-year span. It’s reasonable to ask ourselves whether a class of papers that averages only one mention in future human-subjects research per decade needs to be reconsidered. Even the high end of the range (five citations in future clinical research over a decade) in that surgical study  represents a paltry impact, indeed.
And this isn’t the end of the story. Other research has found that basic elements of study quality like blinding, randomization, and sample-size calculations—which should be readily possible to implement in animal research—are not done in most cardiovascular research papers involving animals, and this problem has not improved over time . Such fundamental methodological shortcomings threaten the validity of conclusions drawn, inhibit reproducibility (a front-of-mind issue even to the public ), and limit the ability of animals’ lives lost to translate into better health for humans. These problems likely are the source of the serious observed discordance between results of animal research and findings of human studies; animal models have grossly and consistently overestimated apparent treatment benefits .
All of that calls into question whether these kinds of studies are, in general, informative. While some may be—as may be the case for some research about bisphosphonates —it appears that most are not. Does this change the ethics of the endeavor? I wouldn’t generalize, but a thoughtful, though retrospective, harm-benefit analysis tried to. That work was performed by a panel that included experts in animal welfare science and veterinary surgery, and it found that less than 7% of studies in the systematic review I mentioned earlier  would be considered ethically permissible using a Bateson’s Cube analysis (a three-dimensional rubric whose axes are animal suffering, the likelihood of benefit to humans, and the importance of research as determined by subsequent impact through citations) . Their study included orthopaedically relevant research, as did another assessment of the validity of animal models for sports-related head injury, which concluded: “Most existing animal models of TBI are not effective for the study of sports-related head impact” . That paper had over 130 references. How many animals do you think were sacrificed in those source studies to reach that not-so-surprising (but nonetheless dismaying) conclusion?
There are other benefits to animal research that these studies don’t quantify; education of trainees may be one. Having done animal research as a student—it seems like a lifetime ago—I did learn from it. I’m just not sure I can make the case that this justifies the taking of animal life, and upon reflection, I’m not sure what exactly it teaches our trainees about our humanity, in either sense of the term: Animal research seems not to tell us much about human disease, and doing animal research (even for the benefit of trainees) that never gets cited seems just plain inhumane. I am happy that in the United States, live animals no longer are used for medical education (as they were when I was a student), though this is a shockingly recent development in medical-school curricula .
Without question, some animal research is needed, and CORR is interested in publishing meaningful, robust research of all kinds, including research that involves animals. But it seems to me that if one is considering starting a project that will use animals’ lives in the course of the study, it’s important to ask oneself some hard questions: Is this work really as important as I think it is? Does it justify the taking of animal lives? How likely is it to pave the way to meaningfully better health for people? And as one answers these questions, it seems important to recognize that many—perhaps most—of those who’ve asked themselves those questions about animal studies that have come before, have gotten the answers wrong.
This month’s spotlight paper  does not answer all the questions we have on this important topic, and some surely will ask whether findings in Turkey generalize more broadly. But the Take 5 interview that follows with Dr. Öztürk about this paper is eye opening, and it’s fair to say that his team’s work puts the ball deep in the court of those who believe animal research as it currently is being conducted in our specialty is appropriate.
Take 5 Interview with Alper Öztürk MD, senior author of “Are the Lives of Animals Well-spent in Laboratory Science Research? A Study of Orthopaedic Animal Studies in Turkey”
Seth S. Leopold, MD:Congratulations on this provocative paper. In it, you wrote that your findings—that animal research on musculoskeletal themes in Turkey just wasn’t very impactful in terms of publication and citations—may not generalize to animal research done in other countries. But the evidence base I covered in my Spotlight commentary suggests that the problem your work points to is, in fact, fairly consistent around the world. How do you really see it?
Alper Öztürk, MD: Thank you for highlighting this important topic. The low publication and citation rates (in Turkey and probably around the world) are likely a result of overestimating the ability of these experiments to influence future research or patient care. Many researchers may be unaware of the controversial reliability and the low translational ability of animal experiments.
While the threshold for initiating an animal experiment varies in different parts of the world, some of the papers you mentioned in your Editor’s Spotlight commentary suggest that perhaps the problem is, in fact, quite generalizable [2, 6, 7, 9]. We may get less out of animal research than many researchers might hope or wish.
Dr. Leopold:Considering that most animal research papers presented at your national congress had little scientific impact, what specifically should a scientist do differently, if that scientist believes he or she has an idea worth exploring, but the idea will require animals’ lives to explore it?
Dr. Öztürk: Researchers should review the “3Rs rule” (replacement, reduction, and refinement) before sacrificing animals to explore new ideas. Determining whether the idea is worth exploring through animal suffering and death is a tough decision that demands honesty and humility from researchers. Researchers should examine their experiments critically from perspectives other than their own. They must consider whether the experiment and sacrifice of animals truly has the potential to yield benefits for humans. Once a researcher has performed an animal study, he or she must feel the moral responsibility to disseminate the results. While scientists perceive they have the right to perform experiments on animals, in light of the studies we’ve been discussing [2, 6, 7, 9] and others, this right should be considered anything but absolute. It is perhaps more a display of man’s power over other species than it is a “right” that humans can justify claiming, and with great power comes great responsibility. I believe that because of the lack of impact of most orthopaedic animal models (and probably other subspecialties) on real-world human healthcare, our profession will be severely and justifiably criticized by future scientists for the things we’ve done while wearing white coats.
Dr. Leopold:I’ve been surprised by the wide variation on what is considered humane treatment of animals from institution to institution around the world. This has been to some degree quantified; one that assessed research involving over 27,000 animals found frequent, severe harms to animals, mainly (but not only) involving inadequate use of anesthesia or analgesia . How do you think journals should handle studies that received institutional ethical approval but whose methods trouble reviewers or editors?
Dr. Öztürk: The fact that animals were sacrificed for an experiment should weigh heavily on journals while evaluating such a study. Poor methodology appears common in animal studies ; in light of the potential for these experiments to cause suffering and death to animals, journals should apply strict standards for papers that use animal models. The ARRIVE (Animals in Research: Reporting In Vivo Experiments) guidelines  can guide authors, reviewers, and editors while reporting or evaluating animal studies. Authors are responsible for the proper welfare of the animal subjects they use. Journals that publish animal research, the reviewers, and the editors are also responsible for the animal studies they have published or reviewed, and they should question the authors about animal welfare and the necessity for the study. Journals should not hesitate to reject studies that do not demonstrate that they have ensured the humane treatment of animals.
Dr. Leopold:I am not opposed to animal research, and you wrote that you are not, either. Can you share with readers what the “right kind” of animal research looks like, and how we can make sure more of it gets done (and less of the other kind)?
Dr. Öztürk: The right kind of animal study should start by asking the right question (Aiming to improve human or animals’ lives). The study should have a reasonable rationale to justify the testing of its research questions on animals. Alternative models (computational, human-based, cell-tissue models, and others) should be considered before designing animal studies, and used whenever reasonable. Researchers should be aware that more-reliable results often can be achieved through these alternatives. The expected results of an animal study should provide useful data and should outweigh the harms to the animals used. When the researcher is convinced about the expected outcomes and there is no alternative model to test the idea, an animal study with proper statistics (demonstrating the minimum but still a sufficient number of animals) can be conducted. While performing the study, animals should be treated humanely, and researchers should try to minimize the pain, distress, and suffering. Animal studies designed without a noble purpose (that is, for the purpose of academic promotion or bulking up one’s curriculum vitae) should never be performed.
Dr. Leopold:Medical schools in the United States no longer use live animals during medical education. This development is relatively recent (the last medical school having stopped the practice less than 5 years ago) , and it represents a change in what had been a widely held belief about the benefits of using animals for education. I surmise that this has been replaced at least to some degree by using animals in surgical research that involves trainees; I think this kind of research is seen as having an educational benefit separate from the research inquiry itself, in that it gives trainees some hands-on experience doing actual “live” surgery. How do you factor this in to your assessment of this subject?
Dr. Öztürk: Animal experiments may provide secondary benefits beyond the main purpose of the study such as hands-on training for trainees. The most-commonly used animals (that were used in the studies we evaluated) were rats and rabbits, and the educational benefit of orthopaedic procedures that performed on these animals (such as fixing the femur of a rat or repairing the Achilles tendon of a rabbit) is questionable. Furthermore, even if this experience contributes to the training of trainees, does it make a poorly designed animal study acceptable? There are many orthopaedic courses including hands-on training on cadavers that will, in my view, provide a more-useful experience for participants. I believe researchers need a change in mindset regarding the benefits of using animals for experimentation purposes.
The author acknowledges Clare M. Rimnac PhD for her thoughtful suggestions, which improved this commentary.
1. Angoa-Pérez M, Kane MJ, Briggs DI, Herrera-Mundo N, Viano DC, Kuhn DM. Animal models of sports-related head injury: bridging the gap between pre-clinical research and clinical reality. J Neurochem. 2014;129:916-931.
2. Gelman A. Essay: The experiments are fascinating. But nobody can repeat them. New York Times. Available at: https://www.nytimes.com/2018/11/19/science/science-research-fraud-reproducibility.html
. Accessed May 28, 2020.
3. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improving bioscience research reporting: The ARRIVE guidelines for reporting animal research. PLoS Biol. 2010;8:e1000412.
4. McVey JS. Last remaining medical school to use live animals for training makes switch to human-relevant methods modern medical education sees no need for animal subjects. Available at https://www.pcrm.org/news/news-releases/last-remaining-medical-school-use-live-animals-training-makes-switch-human
. Accessed May 28, 2020.
5. Öztürk A, Ersan Ö. Are the lives of animals well-spent in laboratory science research? A study of orthopaedic animal studies in Turkey. Clin Orthop Relat Res. [Published online ahead of print May 26, 2020]. DOI: 10.1097/CORR.0000000000001335
6. Perel P, Roberts I, Sena E, Wheble P, Briscoe C, Sandercock P, Macleod M, Mignini L, Jayaram P, Khan KS. Comparison of treatment effects between animal experiments and clinical trials: Systematic review. BMJ. 2007;334:197.
7. Pound P, Nicol CJ. Retrospective harm benefit analysis of pre-clinical animal research for six treatment interventions. PLoS One. 2018;13:e0193758.
8. Ramirez FD, Motazedian P, Jung RG, Di Santo P, MacDonald ZD, Moreland R, Simard T, Clancy AA, Russo JJ, Welch VA, Wells GA, Hibbert B. Methodological rigor in preclinical cardiovascular studies. Targets to enhance reproducibility and promote research translation. Circ Res. 2017;120:1916-1926.
9. Ruan Y, Robinson NB, Khan FM, Hameed I, Rahouma M, Naik A, Oakley CT, Rong L, Girardi LN, Gaudino M. The translation of surgical animal models to human clinical research: A Cross-sectional study. Int J Surg. 2020;77:25-29.