Help Us Strengthen Rigor of Animal Research: Public Feedback Requested


Ever figured out a clever solution to a vexing challenge that affected the rigor of your work with laboratory animals, and then thought that those solutions could improve the quality and transparency of animal research supported across NIH? Recently found yourself at virtual lab meetings brainstorming ways to facilitate translating the findings from your animal study to human biology and disease? Questioned the status quo on how the research culture drives the choice of animal models and the design of experiments? Well, we want to know more.

We recently released a Request for Information (RFI) aimed at enhancing rigor, transparency, and translatability to improve biomedical and behavioral research involving animal models  (NOT-OD-20-130). Your feedback is welcomed electronically here by July 31, 2020.  Note: the deadline to submit comments is expected to be extended to August 21, 2020.

Research involving animal models are extremely valuable and critical for testing hypotheses about the mechanisms by which various biological factors impact the etiology and course of health and disease. To get us there, NIH seeks to achieve “reproducibility through rigor and transparency.”

We have taken several steps to underscore the importance of these precepts, but concerns persist around suboptimal study design, selection of appropriate models, and rigor of statistical design and analyses (see the 2014 National Academies workshop here for more). Moreover, as NIH support for model organism research has remained relatively stable over time (as noted in this 2016 post), addressing these issues is key to proper stewardship of taxpayer funds. And, as my colleague, Dr. Carrie Wolinetz, wrote in her blog in February following a related workshop on non-human primate models, “rigor and transparency necessitates data sharing yet there are challenges…[but] rigorous and reproducible science is part of a virtuous learning cycle.”

A working group of the Advisory Committee to the NIH Director, of which I am a member, is leading this RFI effort. They presented some preliminary thoughts at the June 2020 meeting on how NIH can improve the scientific rigor, reproducibility, translatability, and transparency of the research it supports.  For instance, they discussed the benefits and burdens of preregistering certain animal studies, potential financial implications for grants, and training needs to ensure animal studies are rigorous and transparent. To get at the importance of selecting appropriate models, discussion has focused around ways to optimize the relevance of the studies to human biology and disease. And, they noted we should look at the underlying state of the science and research culture to better understand what incentives or disincentives influence research using animals along with the importance of educating all levels of scientists in good research practice.

So, turning back to the RFI, the working group will consider your comments as they develop their final report (expected in December). We look forward to hearing from you and your colleagues on ways we can strengthen the rigor of NIH-supported animal research, further assuring that future NIH supported activities contribute to our understanding of the complexities and the spectrum of the diseases and disabilities that affect human health.


  1. An excellent platform for providing constructive feedback in NIH USA-sponsored ethical animal research with eventual translational research and public health impact in the ever-expanding healthcare research with timeline-based strategic priorities!
    The crisp snapshot is indeed enlightening with special emphasis on research involving animal models that are extremely valuable and immensely critical for testing scientific rationale-based sound research hypotheses relating to underlying celullar/molecular/genetic networks/mechanisms by which various biological and/or environmental factors/determinants impact the overall etiology and course of health and disease on a global platform.
    In this context, it is essential to draft a policy-driven framework with consensus from experts involved in biomedical/translational/public health research with an eventual clinical impact so as to translate the rigorous animal models of disease related research outcomes into patient-friendly ethical, transparent and reproducible human subject research.
    Animal research training with emphasis on ethical animal research should be developed with module-based course-works so that animal-care and ethical guidelines are stringently followed during the course of animal research and in case of untimely death of animals involved in research viz. animal rearing, feeding animal-specific diets for various experimental protocols, drug-research related euthanasia an/or animal-organ-ligation-modeling experiments, the euthanized animal(s) and/or dead animal(s) should be discarded as per the stringent NIH USA-specified protocols and Instititional Review Board guidelines and the biohazard waste-disposal team should be accordingly notified with due record of the specific strain, gender (female vs male) of the animal model(s) so that a formal register of animal-experimentation with experimental protocols including the reason for animal’s untimely death dueing the experimental procedure be maintained in a specific repositoey at the parent organization/laboratory with due coding of the animal(s) euthanized/sacrificed in experiments.
    The NIH USA should be credited and acknowledged for federal grants/fund support for ethical animal research that may be further extrapolated in human subject research.
    As an elegant example, animal models of fertility (female vs male infertility) and/or pregnancy/obstetrics-gynecology research should be utilized in scientically sound experimental research by senior investigators with proven publication track records as senior/lead authors as evident in high impact leading scientific publications to ensure that the animal models in research are aptly and ethically valuable models in fruitful research with eventual meaningful outcomes in reproductive medicine. Porcine/swine, equine, murine, dogs/bitches, cats, etc models in infertility and accordinglyrelevant animal models in other diseases (cancers, cardiovascular, airway inflammation/asthma, sepsis, neurogeneration, liver diseases, etc.) should prove immensely valuable for developing disease-specific interventions/innovative treatment modalities, immunotherapy regimens and pharmacological scaffolds/innovative drugs for research and development in the years to come.
    Overall, I gained critical research insights in animal research and NIH USA-sponsored ethical biomedical research for future professional scientific pursuits!

  2. I agree that rigor and reproducibility are essential qualities of research. Please do not use an indiscriminate approach to impede research in humans based on concerns about work in animals/cell lines/etc. For work in humans, requiring scientific journals to permit (at least online) reporting of adequate detail in methodologies and results would be more fruitful. It is difficult to find a journal that will accept a report with adequate detail from a study, without dividing results into several manuscripts that will not show the whole picture. This may be due to publication costs but these days there are ways to accommodate it.

  3. Killing the “mouse TCGA” proposal was one of the most detrimental decisions NCI made in this regard. Nothing would impact reproducibility and rigor for cancer research better than having a common reference -omics database for animal models of cancer. Model selection, study design, and statistical expectations would all experience associated improvements, and our work as scientists would be made more efficient rather than less.

  4. One of the NCI’s most negative decisions in this regard was to abandon the plan for the “Mouse TCGA.” There will be nothing more impactful than providing a popular omics database for animal cancer models, on reproductivity and rigor for research on cancer. Model selection, study design, and statistical expectations would all be made more efficient rather than less efficient by all associated improvements.

    Thanks for such a nice blog .

Before submitting your comment, please review our blog comment policies.

Leave a Reply

Your email address will not be published. Required fields are marked *