Ch. 5— The Use of Animals in Research • 107 
Table 5-5.— Percentage of Papers (Average, 1980-83) 
Using Animal, Nonanimal, and Human Subjects in 
15 Biomedical and Behavioral Research Journals 
Surveyed by OTA 
Percentage of papers using: 
Journal 
Animals 
Nonanimals 
Humans 
Biomedical research: 
American Journal of 
Veterinary Research .... 
96 
4 a 
0 
Journal of Animal 
Science 
96 
4 b 
0 
Endocrinology 
91 
8 
1 
American Journal of 
Physiology 
90 
6 
4 
Anatomical Record 
88 
3 
9 
Proceedings of the Society 
of Experimental Biology 
and Medicine 
83 
10 
7 
Journal of Immunology . . . 
71 
10 
19 
Journal of the National 
Cancer Institute 
60 
9 
31 
Developmental Biology . . . 
54 
46 
0 
Journal of Biological 
Chemistry 
39 
52 
9 
Cell 
31 
67 
2 
American Journal of 
Cardiology 
12 
0 
88 
Behavioral research: 
Behavioral and Neural 
Biology 
96 
4 
0 
Journal of Comparative 
and Physiological 
Psychology 
96 
3 
1 
Physiology and Behavior . . 
93 
1 
6 
a Primarily virus research. 
^Primarily computer modeling or grain fermentation applicable to ruminant 
nutrition. 
SOURCE: Office of Technology Assessment. 
tions . The conservative scoring procedure tended 
to underestimate the use of alternatives to ani- 
mals as defined in this assessment (see ch. 2). For 
instance, if an experimental protocol used both 
animal and nonanimal methods, it was catego- 
rized under use of animals. If a study involved 
both nonanimal methods and humans, it was 
counted as use of humans. Further, if a study in- 
volved both animal methods and humans, it was 
counted as use of animals. The approach used to 
categorize protocols took into consideration only 
the replacement, not the reduction or refinement, 
of animal methods. Whether a protocol involved 
1 or 100 animals, it still fell under the category 
of “use of animals,” and all reports bore equal 
weight in determining percentage of protocols 
using humans or animal or nonanimal methods. 
In addition, there was no attempt to quantify the 
pain or stress of an animal in an experiment or 
to distinguish between different vertebrate spe- 
cies. Any alternative protocol, therefore, that 
tended to reduce or refine an existing animal pro- 
cedure was still categorized as "use of animals.” 
An example of the problem of overestimation 
of animal use by a survey such as this exists in 
immunology. Today, antibodies, needed in most 
immunology research, can be obtained by inject- 
ing rabbits with foreign proteins, or antigens, and 
extracting the antibodies that the rabbit produced 
or by using mouse spleens and the monoclonal 
antibody technique to produce antibodies to an 
antigen. Each process requires animals. Once the 
monoclonal cells are in culture, however, there 
is a virtually unlimited supply of the needed anti- 
body, and there is essentially no further need for 
animals. Thus the monoclonal technique can de- 
crease animal use, as was the case in many of the 
most recent articles surveyed in the Journal of 
Immunology. These articles, though, were still 
coded under the “use of animals” category be- 
cause the primary methods and materials in- 
volved animals. But the total number of animals 
in a given experiment decreased, for they were 
used in just one aspect of the experiment instead 
of two. The monoclonal antibody technique is be- 
ing used as an alternative to the repeated use of 
rabbits, yet its impact is underestimated in a sur- 
vey such as this. The OTA scoring of protocols 
published in the Journal of Immunology did not 
reflect certain reductions that are currently be- 
ing implemented. 
Along with underestimating the implementation 
of nonanimal methods, the boundaries within 
which the OTA survey was carried out also 
tended to overestimate the use of animals as ex- 
perimental subjects. This was due principally to 
two factors included in the scoring procedure 
under animal use— epidemiologic studies and the 
study of biological molecules obtained from 
animals. 
Epidemiologic data are the primary sources in 
some articles in the American Journal of Cardi- 
ology, the Journal of the National Cancer Insti- 
tute, and in many veterinary studies. These pro- 
tocols were included under “use of animals,” vet 
they did not manipulate animals in anv way as 
experimental subjects. 
