12 • Alternatives to Animal Use in Research, Testing, and Education 
and reliability of the data. A registry along these 
lines would probably be 3 to 15 times as complex 
and costly as the unsuccessful LADB. 
As alternative methods are developed and im- 
plemented, a computerized registry of informa- 
tion about these novel techniques might serve to 
speed their adoption. In 1985, the NLM incorpo- 
rated “animal testing alternatives” as a subject head- 
ing in its catalogs and databases, which help users 
throughout the world find biomedical books, arti- 
cles, and audiovisual materials. In amending the 
Animal Welfare Act in 1985, Congress directed the 
National Agricultural Library to establish a serv- 
ice providing information on improved methods 
of animal experimentation, including methods that 
could reduce or replace animal use and minimize 
pain and distress to animals. 
ECONOMIC CONSIDERATIONS 
The total dollar cost of the acquisition and main- 
tenance of laboratory animals is directly related 
to the length of time animals stay in the labora- 
tory. With no accurate source of data on various 
species’ length of stay, it is impossible to calculate 
the actual total dollar cost of animal use. Reduc- 
ing the number of animals used can lower acqui- 
sition and maintenance costs. Yet, the overall sav- 
ings will not be proportionate to the smaller 
number of animals used, as the overhead costs of 
breeding and laboratory animal facilities must still 
be met. 
Animal use carries with it both great expense 
and major economic and health benefits (see 
chs. 5, 7, and 11). Nonetheless, it is difficult to ex- 
press many of the costs and benefits monetarily. 
What price does society put on the pain and dis- 
tress of an animal used in research, for example, 
or on the life of a person saved by a new medical 
treatment that was made possible by the use of 
animals? 
In research, there is no way of knowing when 
a particular result would have been obtained if 
an experiment had not been done. Thus, it is im- 
possible to predict many of the costs related to 
the use of alternatives in research. Attempts to 
do so are likely to result in economic predictions 
with little basis in fact. 
The primary reason a company conducts ani- 
mal tests is to meet its responsibilities to make safe 
products under safe conditions. For pharmaceu- 
ticals, the need extends to the assurance of product 
effectiveness. In testing, animal methods gener- 
ally are more labor-intensive and time-consuming 
than nonanimal methods, due to the need, for ex- 
ample, to observe animals for toxic effects over 
lifetimes or generations. Testing can cause delays 
in marketing new products, including drugs and 
pesticides, and thus defer a company’s revenue. 
Rapid, inexpensive toxicity tests could yield ma- 
jor benefits to public health. There are more than 
50,000 chemicals on the market, and 500 to 1,000 
new ones are added each year. Not all must be 
tested, but toxicologists must expand their knowl- 
edge of toxic properties of commercial chemicals 
if human health is to be protected to the extent 
the public desires. Rapid and economical testing 
would facilitate the expansion of that knowledge. 
Government regulatory practices can be read 
as promoting animal testing, although the laws 
and practices appear flexible enough to accept 
alternatives when such tests become scientifi- 
cally acceptable. To date, regulatory practices 
have not, in fact, provided a basis for companies 
to expect that acceptance of alternative methods 
will be an expedient process. In addition to re- 
sponding to regulatory requirements, companies 
conduct animal tests to protect themselves from 
product liability suits. Here, the necessary tests 
can exceed government requirements. 
Because of the great expense and long time re- 
quired for animal research and testing, priority 
in research results has considerable value to in- 
vestigators and testing results bear considerable 
proprietary value for industry. Some data are made 
public by statute, and various arrangements can 
be made for sharing testing costs. Yet many data 
are held in confidence, for example, by the com- 
pany that generated them. 
