236 • Alternatives to Animal Use in Research, Testing, and Education 
User Friendliness and User’s Needs 
The interactive software used in LADB was de- 
signed in 1975. As such, it predated many major 
software developments that have emphasized "user 
friendliness.” The users the system was aimed at— 
biologists— found it hard to retain procedural 
familiarity with infrequent use. 
Another problem with the use of the LADB data, 
according to the FASEB ad hoc panel (2), was the 
inability to perform on-line statistical comparisons 
between different data sets. This limitation, which 
makes some desirable statistical comparisons dif- 
ficult to perform, arose from inadequate design 
and would probably not be a problem with today’s 
software. 
User Community 
LADB was publicly available for too short a time 
to permit many conclusions to be drawn about 
the users. By definition of the content, its users 
would be expected to be pharmacologists and toxi- 
cologists concerned with toxicity testing, particu- 
larly chronic toxicity testing. This community, 
numbering about 3,000 to 5,000 scientists, is far 
smaller than the community of basic biological 
scientists (about 200,000). The pool of prospective 
users of LADB, therefore, seems too small to sus- 
tain it. 
Peer Review of LADB Design and Data 
Although the Institute of Laboratory Animal Re- 
sources of the National Academy of Sciences evalu- 
ated LADB in 1 9 78, it had not been involved in the 
original design considerations. Similarly, FASEB 
entered the review process in 1979— too late to 
have substantial impact on the design and most 
of the file -building process. The March 19 80 FASEB 
review (2) pointed out several major design prob- 
lems, including lack of on-line availability of the 
individual animal data files . The LADB records that 
are searchable on-line are composites from groups 
of animals. Failure to include data on individual 
animals prevents users from performing statisti- 
cal comparisons between different data sets. 
Lessons Learned From LADB 
The acceptance of a biology data bank by the 
user community and its success in supplying use- 
ful research and testing data are actually deter- 
mined well in advance of the collection or dissem- 
ination of data. The first step in assembling a 
computerized data registry should be the clear def- 
inition of its potential users and their specific needs . 
No adequate study of this nature was performed 
prior to the original design of LADB. The results 
of a preliminary feasibility study should identify 
the various users , their needs , and their desire (or 
lack thereof) to use and support the proposed data- 
base (3). 
A 1981 FASEB report, “Guidelines for Develop- 
ment of Biology Data Banks” (4), emphasized three 
important steps in planning and developing a data 
bank of biological information. First, the stimulus 
for establishing a research and testing data bank 
may be the realization by a scientist, a government 
agency, or a private organization that the required 
information is not readily accessible from pub- 
lished, unpublished, or on-line resources. Never- 
theless, the need for such an information resource 
must be determined independently. Most appro- 
priately, this is done by an organization unrelated 
to the proposing institution. Determination of need 
involves answering the following questions: 
• How many institutional, organizational, or in- 
dividual users would find the database useful? 
• How many would be willing to subscribe, and 
to what extent would cost be a factor in sub- 
scribing? 
• How many institutions, organizations, or in- 
dividual scientists could supply data? How 
many would? 
• How are potential users distributed among 
disciplines? 
• How much unpublished and presently inac- 
cessible data could be made available to in- 
vestigators by developing a data bank? 
If the responses to these questions indicate a solid 
foundation of perceived need, then the establish- 
ment of the data bank is probably justified. 
