### General Information

Federal laws and Institutional policies require us to apply the principles of the 3Rs (replacement, reduction, and refinement) to animal research. U.S. Government Principle #3 states “The animals selected for a procedure should be of an appropriate species and quality and the minimum number required to obtain valid results.” PHS policy states that the institution provide “training or instruction in research or testing methods that minimize the number of animals required to obtain valid results and minimize animal distress, offered to scientists, animal technicians, and other personnel involved in animal care, treatment, or use.” The Animal Welfare Regulations state that the institution must include training in “The concept, availability, and use of research or testing methods that limit the use of animals or minimize animal distress.”

Reduction is desirable in any procedure, however directly humane, which employs large numbers of animals. Reduction is of great importance and, of all modes of progress on the three Rs, it is the one most obviously, immediately, and universally advantageous in terms of efficiency. One general way in which great reduction may occur is by the right choice of strategies in the planning and performance of whole lines of research. Many investigators settle early in their research career on some strategy that appears to suit them and are liable never to change from that strategy.

The term **Reduction Alternatives** describes methods for obtaining comparable levels of scientific procedures or for obtaining more information from a given number of animals. Proper statistical design is essential. The precision of an experiment depends mainly on the sample size and error variance. Careful attention must be given to the type of endpoint used. Qualitative endpoints (e.g., dead/alive) often involve severe pain and distress and generally provide less information than quantitative measurements. There is evidence that poor experimental design and inappropriate statistical analysis has lead to the inefficient use of animals. This may be due to a low level of statistical expertise in the investigators so that they are unaware of the potential value of obtaining statistical advise. Alternatively, they may be unable to do so because of a lack of qualified biometricians with experience in their field of interest. In a research project by Hume (1957), he showed that every time any particle of statistical method is properly used, fewer animals are employed than would otherwise have been necessary.

**Common Methods to Reduce Animal Numbers**

### Power Analysis

Statistical significance and biological significance are not the same thing; however, statistical and biological significance can be linked through the use of statistical power analysis. The statistical power of a test is the probability of getting a statistically significant result, given that the null hypothesis is false. Power is proportional to the sample size, significance criterion and effect size, and is inversely proportional to the variance in the population. Effect size is a measure of biological significance: it is the difference between the results predicted by the null hypothesis and the actual state of the population being tested. Power analysis can be used to determine whether the experiment has a good chance of producing a statistically significant result if a biologically significant difference exists in the population. Or, in other words, whether the experiment has a high power, given a biologically significant effect size. What constitutes “high power” is best left to the researcher, but conventions of 0.8 and 0.95 have been suggested in the literature. It is beyond the scope of this general introduction to teach statistics to the researchers, but there are methods that are out there that are available. Consultation with a biostatistician can yield benefits to the PI and the experimental animals.

### Pilot Studies

Pilot studies are a good way to reduce the number of animal used; the IACUC may even require a pilot study when reviewing a protocol. Pilot studies can be used to estimate variability and evaluate procedures and effects.

### Appropriate Use of Endpoints

The precision of an experiment depends mainly on the sample size and error variance. Careful attention must be given to the type of endpoint used. Qualitative endpoints (e.g., dead/alive) often involve severe pain and distress and generally provide less information than quantitative measurements. More information can generally be found using quantitative endpoints and can, in some instances, lead to a reduction in the number of animals used during an experiment.

### Sharing Animals

In some instances, it is possible to share research animals. For instance, animals euthanized by one investigator can provide tissue for use by another investigator. There are instances in which this should never be attempted (e.g., animals have been exposed to biological hazards, chemical hazards, and/or recombinant DNA), but it is a method to reduce animal numbers that should be explored by researchers.

### Use Quality Animals and Veterinary Care

When PIs use the correct choice of an animal model- one that uses healthy, genetically similar animals- it generally decreases variability and, hence, animal numbers. You can minimize the loss of animals by providing good post-operative care, avoiding unintended breeding, and planning ahead so that the appropriate number of animals needed for the studies are ordered and/or bred.

### Computer Simulations

Though not always possible, there are cases in which there are computer simulations available which can mimic functions of physiology. These are typically most helpful in the case of training protocols.

### Use of Cell Culture

When possible, consider the use of cell cultures rather than animals. For example, there are *in vitro* systems which use cell culture to generate monoclonal antibodies rather than using laboratory animals.

### Auto Controls

It would be helpful, whenever possible, to design experiments in which animals serve as their own control. For example, if a procedure were to be performed on a limb, instead of doing the procedure on both limbs of an animal- and having separate control animals- do a unilateral procedure with the control being the opposite limb. Though this is not always possible, it is a well-established method for reducing animal numbers.

### New Instrumentation and Techniques

Using new instrumentation or innovative techniques that can improve precision can reduce the number of animals needed for a study. This has the added benefit of also being a refinement technique for the protocol.

### Appropriate Experimental Design

Careful experimental design by appropriate choice of control groups and standardizing procedures to minimize variability; for example, University of Leicester researchers designed experiments to reduce the number of mice required to investigate treatment combinations in a study of tumors resulting in an average of 1.5 mice per treatment combination instead of the 20 to 40 mice per treatment combination usually used.

### Helpful Links

- A Simple, Rapid and Reliable Method for Selecting or Assessing the Number of Replicates for Animal Experiments
- IACUC- non-statistical approach to calculating appropriate animal numbers (PDF)
- Isogenic.info
- Rice Virtual Lab in Statistics
- Statistical Considerations for a Parallel Trial where the Outcome is a Measurement
- Web Interface for Statistical Education
- Power and Sample Size