4 • Alternatives to Animal Use in Research, Testing, and Education 
DEFINITION OF TERMS 
In this report, animal is defined as any non- 
human member of the five classes of verte- 
brates: mammals, birds, reptiles, amphibians, 
and fish (see ch. 2). Within this group, two kinds 
of animals can be distinguished— warm-blooded 
animals (mammals and birds) and cold-blooded ani- 
mals (reptiles, amphibians, and fish). Other crea- 
tures customarily included in the animal kingdom, 
such as invertebrates (e.g., worms, insects, and 
crustaceans), are excluded by this definition. The 
use of human subjects is not examined in this 
assessment. 
The concept of alternatives to animal use has 
come to mean more than merely a one-to-one 
substitution of nonanimal methods for animal tech- 
niques. For alternatives, OTA has chosen a def- 
inition characterized by the three Rs: replace- 
ment, reduction, and refinement. 
Scientists may replace methods that use animals 
with those that do not. For example, veterinary 
students may use a canine cardiopulmonary -resus- 
citation simulator, Resusci-Dog, instead of living 
dogs. Cell cultures may replace mice and rats that 
are fed new products to discover substances poi- 
sonous to humans. In addition, using the preced- 
ing definition of animal, an invertebrate (e.g., a 
horseshoe crab) could replace a vertebrate (e.g., 
a rabbit) in a testing protocol. 
Reduction refers to the use of fewer animals. 
For instance, changing practices allow toxicolo- 
gists to estimate the lethal dose of a chemical with 
as few as one-tenth the number of animals used 
in traditional tests. In biomedical research, long- 
lived animals, such as primates, may be shared, 
assuming sequential protocols are not deemed in- 
humane or scientifically conflicting. Designing ex- 
perimental protocols with appropriate attention 
to statistical inference can lead to decreases (or 
to increases) in the numbers of animals used. Or 
several tissues may be simultaneously taken from 
a single animal as a result of coordination among 
investigators . Reduction can also refer to the mini- 
mization of any unintentionally duplicative exper- 
Resusci-Dog, Canine Cardiopulmonary- 
Resuscitation Simulator 
Photo credit: Charles Ft. Short, 
New York State College of Veterinary Medicine, 
Cornell University 
Resusci-Dog, aplastic mannequin linked to acomputer, 
can simulate an arterial pulse, and pressure can be 
applied to its rib cage for cardiac massage or cardio- 
pulmonary resuscitation. Resusci-Dog has replaced 
about 100 dogs per year in the training of veterinary 
students at the New York State College 
of Veterinary Medicine. 
iments, perhaps through improvements in infor- 
mation resources. 
Existing procedures may be refined so that ani- 
mals are subjected to less pain and distress . Refine- 
ments include administration of anesthetics to ani- 
mals undergoing otherwise painful procedures; 
administration of tranquilizers for distress; hu- 
mane destruction prior to recovery from surgical 
anesthesia; and careful scrutiny of behavioral in- 
dices of pain or distress, followed by cessation of 
the procedure or the use of appropriate analgesics. 
Refinements also include the enhanced use of non- 
invasive imaging technologies that allow earlier 
detection of tumors, organ deterioration, or meta- 
bolic changes and the subsequent early euthana- 
sia of test animals. 
Pain is defined as discomfort resulting from in- 
jury or disease, while distress results from pain, 
anxiety, or fear. Pain may also be psychosomatic, 
resulting from emotional distress. Although these 
are subjective phenomena, pain and distress can 
