(5) 
Recommendations : 
1. EPA, OSHA, FDA, and CEQ should have full voting membership 
on the RAC. This is essential now that the NIH is accepting volun- 
tary compliance by industry. In order to protect proprietary in- 
formation, the RAC will have to^ hal'd meetings not open to the public. 
We need to have some assurance that our concerns will be represented. 
Perhaps more importantly, the RAC could use the advice of experi- 
enced regulators, particularly as they begin to deal with industry. 
The Interagency Committee is not adequate participation and 
input for these agencies. It has not met once this year. The 
advice of EPA and OSHA, in particular, will always be "appropriate" 
(p. 33051), they should be fully involved in the decision-making. 
2. The majority of the open spaces on the RAC should be filled 
with individuals nominated directly from the public. Until we have 
members who share our perspective, and whose judgement we trust, 
the decisions of the RAC will always be suspect. This trust is 
in the long-term interests of the NIH, as well as ourselves. 
Penalties 
HEW must find some credible mechanism for ensuring enforcement 
of the guidelines; withdrawal of NIH funding is totally inadequate 
The sanction is so severe that the NIH will probably never use it, 
except for the most gross violations, witness the recent experience 
with UCSF and Harvard. 
HEW must find some mechanism by which penalties can be assessed 
appropriate to the infraction, an then use them as required. 
IV. Responsibilities of the Institution 
Health Monitoring 
Institutions are only required- to- determine the necessity of 
doing health monitoring, not to actually do it. There is nothing 
in the guidelines that requires medical monitoring, data collec- 
tion, maintenance of records, or most importantly good experiments 
designed to determine what if anything is happening with recombin- 
ant organisms inside human bodies. 
Very little study has been done on the epidemiology of recom- 
binants in humans. And although we've heard repeatedly that no 
one has been injured by this kind of research, they're still saying 
that about commercial nuclear reactors. It is important to note 
that the most dangerous kinds of experiments have not been done . 
Almost all the work has been executed with the safer recombinants, 
primarily because of the lack of high containment labs. But with 
the new guidelines, what used to be done in P-4 and P-3, can now 
go on in P-3 and P-2. New host-vector systems, work in wild-type 
e. coli now allowed through exemption, and large scale processes 
will make what little we know essentially inapplicable. 
[ 356 ] 
