108 • Alternatives to Animal Use in Research, Testing, and Education 
Obtaining many biological molecules that are 
studied experimentally requires that they be ex- 
tracted from animals who produce them. Subse- 
quent experiments on these molecules themselves 
(reported, for example, in the Journal of Biologi- 
cal Chemistry) do not involve animals at all. In 
such cases, animals may be used in preparation 
for an experiment but are not actually involved 
in the experiment being performed. Therefore, 
protocols that involved animals as donors of bio- 
logical molecules (e.g., bodily fluids) for an exper- 
iment prior to its initiation were also included 
under use of animals, and this tended to overesti- 
mate the use of animals as research subjects. 
It is important to distinguish between the num- 
ber of published articles involving animal meth- 
ods and the actual number of animals used in re- 
search. The OTA survey provides no information 
on the latter. Some protocols may involve only 
a few animals, while others may employ tens or 
hundreds. Moreover, depending on the species 
and type of research, some subjects might be used 
in multiple experiments. In primate research, for 
example, it is not uncommon for animals to be 
used in a succession of either related or unrelated 
studies over a period of years; this would not be 
the case for rodents. 
SUMMARY AND CONCLUSIONS 
Biomedical and behavioral research center on 
the understanding of human health and disease 
and rely on animals to achieve this goal. They use 
animal subjects to understand not only human 
phenomena, but animal phenomena as well. The 
broad spectrum of enterprises involved in these 
fields of research includes disciplines ranging 
from anatomy to zoology. Although the varied dis- 
ciplines that make up biomedical and behavioral 
research have distinct foci, they often overlap. 
Animals are used throughout these disciplines 
to address an array of questions. Nonhuman pri- 
mates, for example, have contributed to an under- 
standing of polio, hepatitis B, high blood pressure, 
Parkinson’s disease, baldness, menopausal hot 
flashes, and other human conditions. Beyond the 
nonhuman primates, diverse species are used in 
biomedical research because of their anatomical, 
physiological, and metabolic similarities to or 
differences from humans. Principles and tech- 
niques developed in varied animal species (e.g., 
dog, horse, and sheep) may combine to support 
a single application to humans, as in the case of 
coronary artery bypass graft surgery. In be- 
havioral research, different animal species may 
also be used to learn about characteristics unique 
to the species under study, usually one of eco- 
nomic importance or intrinsic interest to humans. 
Animals may suffer pain or distress in the 
course of research on the mechanism of pain, or, 
more generally, as a byproduct of experimental 
procedures. In such cases, the investigator is 
obliged to supply pain relief to the animal or to 
justify withholding pain-relieving drugs as nec- 
essary to the experiment. Institutional animal care 
and use committees play an important role in 
overseeing this process (see ch. 15). Pain relief is 
usually effected by the administration of analge- 
sic, anesthetic, or tranquilizing agents. Indices of 
pain can usually be recognized in experimental 
animals, and experimental procedures can be 
ranked according to estimates of the degree of 
pain produced. Such a ranking provides a basis 
for efforts to minimize the pain caused by re- 
search procedures. 
An OTA survey of published research reports 
in 15 scientific journals documented the preva- 
lence of animal v. nonanimal protocols in contem- 
porary research. Each research journal, and per-, 
haps the discipline it represents, can be identified 
by a characteristic balance of protocols using ani- 
mals, nonanimals, and humans. The data permit 
research journals to be classified as representing 
disciplines that either rely on nonanimal meth- 
ods or that do not incorporate such methods (or 
do not have them available). In the same way that 
research itself is not monolithic, patterns of ani- 
mal use and the use of nonanimal methods among 
research disciplines are not uniform. 
