IACUC Policy 005: Justification of Animal Numbers

1. Purpose & Scope: The IACUC must ensure all animals described in an animal protocol are accounted for and justified. This mandate is found in the Guide for the Care and Use of Laboratory Animals (8th Edition, NRC 2011, p25) , wherein protocol review includes “justification of the species and number of animals proposed” and further, “whenever possible, the number of animals and experimental group sizes should be statistically justified (e.g., provision of a power analysis)”. In addition, for USDA-covered species, numbers of animals to be used must be rationalized (CFR9, Part 2, Subpart C, §2.31(d)8(e)(2)).

• The reason for animal number justification is to potentially Reduce the animal numbers requested, Refine the experimental endpoints to minimize pain and suffering, and Replace animal usage with alternative scientific models (Three Rs of animal research).
• It may often be appropriate to request additional animals to account for possible animal losses, experimental failures, training in new or difficult technical procedures, or other unforeseen needs, but the reasons for these additional animals must be clearly explained to IACUC reviewers.

2. Policy:

2.1 Studies in Which Statistical Justification is Not Possible:

• Pilot studies: Animals may be needed for pilot or proof-of-concept studies. Animal numbers for such studies may be justified based on the probability of observing a desired effect of the experimental procedure or to evaluate a new paradigm. These studies should require only small number of animals enabling collection of data that can then support a statistical sample size calculation for future experiments.
• Breeding protocols may be required to maintain strains of animals available for future research activities. Instead of statistical justification, the PI should provide information about the number of animals required to maintain the animal strain. Protocols that include both breeding and research components may be required to provide statistical justification for the research objective in addition to justification of the numbers of animals required to maintain the strain.  This may be based off of other protocols where animals are transferred to for experimentation.
• For teaching protocols, the number of animals requested is usually based on the number of students in the class. Instead of statistical group size justification, the PI should justify the animal numbers in terms of the minimum number of animals required to meet the specific teaching objectives.
• Non-intervention field studies (e.g., behavioral observations of wild animals in their native habitat) do not require justification of specific animal numbers.  If animals are being handled and samples collected a justification of the required number of animals should be provided.
• Protocols for which animals are required to produce tissues, cells, or cellular sequences or similar biological end products (e.g., in cell culture, microarray, etc.). The PI should work backward in their justifications: Start by justifying the required number/ amount of products, and then explain how you chose the number of animals required to provide that amount.

2.2 Example Of Statistical Justification of Animal Numbers:

“Fourteen (14) rats per group are required based on the following power calculation and our estimation of attrition due to surgical failure. We have conducted similar studies in the past investigating the effects of these lesions on stress-induced weight change. Using the means from this study (31.89 for control/sham, 31.14 for control/lesion, 3.25 for stress/sham and 15.07 for stress lesion with an average standard deviation of 7.397), we calculated an effect size of 1.565. Conducting a power analysis (using GPOWER software) for a one-way ANOVA with this effect size revealed a sample size of 12 per experimental group. This value is consistent with numbers that I have used in the past for similar studies. Because rats will undergo surgery to implant bilateral guide cannulae aimed at a specific anatomic area of the brain, we anticipate that as many as 15% of rats will have one or both cannulae misplaced (outside the desired structure). Thus, we are requested 2 additional rats per group for these experiments. Thus, 14 rats per group are requested for all experiments.

2.3 Example When Statistical Justification Is Not Available:

• Pilot Study: This is a pilot study intended to demonstrate the feasibility and safety of this approach. The number per group is based on the experience of the PI for studies of this nature. Observations from this pilot study will be used to guide formal statistical power calculations for future efficacy studies.
• Teaching Protocol: We plan to have up to six 3-person teams of students per lab section. We plan to have two lab sections for one semester (Spring Semester) each year. We will design these labs to enable each team to observe up to 6 rats per lab period. (When the PI executed this type of 3-hr student lab at a previous institution, he found than when the lab is carefully planned, teams of students can observed up to 6 rats, each for 18-minute observation periods.) Thus we will require 72 rats per Spring Semester when the class is scheduled.
• Cell Yield: The number of animals to be used in each experiment is based on the expected number of different cells which can be isolated from individual animals and has been published by this laboratory [Reference cited.]. The cell populations and anticipated cell yield/mouse from heart are as follows: CD4+ T cells, 6 x 106 cells; and bone marrow cells, 9 x 107 cells.
• “Numbers of mice required/group is determined by the formula: #cells needed for an experiment/# cells which can be obtained per mouse. The experiments should need to be done only once as the statistics will be obtained from replicate samples in tissue culture.  

2.4 Updating Animal Numbers Requested at The 3-Year De Novo review:

At the time of 3-year de novo review, the investigator should specify the total number of animals being requested for the next 3-year period in the “Number Requested” box in the electronic protocol.  This should include animals currently on the protocol which are continuing in long-term studies and any new animals being added.