NOTICES 
33053 
example, some feel that work In a P3 
facility conveys a desirable sense of 
hazard, whereas a reduction to the P2 
level will promote an undesirable re- 
laxation of vigilance, it has also been 
suggested that an increase in the op- 
tions increases the difficulty of control 
implementation of the guidelines, 
e commentators object to specific 
options provided at the P3 and P4 
levels. NIH has been urged to include 
a better explanation of the rationale 
for this flexibility. 
Indeed, the calculus of switching 
physical and biological containment 
levels has been questioned. Does an in- 
crease in biological containment from 
EK1 to EK2 truly compensate a reduc- 
tion in physical containment from P3 
to P2? 
The scale of either form of contain- 
ment from least to greatest is not nec- 
essarily linear, and substitutions are 
only roughly approximate. Neverthe- 
less, there are some numerical bases 
for comparison. 
For example, a class III biological 
safety cabinet is required at the P4 
level (if a positive pressure suit is not 
used); whereas at P3, one can work in 
an open-front biological safety cabi- 
net. The class III cabinet is virtually 
an absolute containment system. It is 
certified gas-tight when tested under 
positive pressure. It is operated under 
negative pressure to gain optimum 
safety. It provides at least a 10,000- to 
100,000-fold increase in safety over 
that provided by a Class I or II cabi- 
net, which is required at the P3 level. 
The relative safety of these two con- 
tainment cabinets is based on the effi- 
ciency of their exhaust-air treatment 
systems. The exhaust-air treatment 
for the class in cabinet is provided by 
two HEPA filters installed in series. 
This arrangement gives a containment 
efficiency of at least 99.99 percent. 
The exhaust-air treatment for class I 
and II cabinets, with only one HEPA 
filter, provides a containment efficien- 
cy of 99.99 percent. The potential for 
escape of microorganisms across the 
open front of the class I and II cabi- 
nets is similar to that for escape 
through the exhaust-air treatment 
system under operating conditions. 
These cabinets must meet a perform- 
ance criterion which permits fewer 
than 20 microorganisms to escape 
through the open front when lx 10 9 
(100,000,000) to 8x10' (800,000,000) 
microorganisms are experimentally re- 
leased within the cabinet. The degree 
of protection provided by the class I or 
II cabinets is equivalent to the in- 
crease in safety at the P3 level over 
that provided at the PI level which 
allows open-bench operations. 
The symbol HV (Host-Vector) is 
used in the PRG-NIH to designate 
biological containment systems encom- 
passing the present EK systems. HV2 
is defined in terms of a probability of 
escape of recombinant DNA of less 
than 1 in 10® (1 in 100.000,000). in con- 
sidering "equivalency” between P and 
EK levels, it is recognized that the two 
systems are conceptually different. 
Biological safety cabinets are designed 
primarily for the protection of the lab- 
oratory worker, and all physical con- 
tainment protection stops at the walls 
of the laboratory. Biological contain- 
ment continues to operate even were 
an organism to escape from the labo- 
ratory. 
The flexibility allowed in alternate 
P and HV levels is carefully explained 
in the text of the PRG-NIH, and the 
investigator must follow the explicit 
requirements set forth in Part III of 
the proposed guidelines and Tables I 
and n. 
Redundancy. A question has been 
raised concerning redundancy in the 
safety systems to insure that alternate 
systems will come into play in case of 
an emergency— for example, power 
failures or major accidents. The con- 
cept of redundancy is inherent in the 
design of the containment systems 
used in recombinant DNA research. 
Redundancy, however, is provided by 
standby systems, but rather by design 
features and operational requirements 
of the safety systems used. For exam- 
ple, primary containment at the P4 
level is provided by the gas-tight class 
III cabinet system. These cabinets are 
also maintained under negative air 
pressure, which would provide protec- 
tion against the release of microrgan- 
isms in the event that a glove were to 
rupture or a leak to develop. Similarly, 
the physical isolation of the class III 
cabinet would not be compromised in 
the event of a power failure. However, 
since the redundant protection pro- 
vided by the negative pressure would 
be compromised, personnel would be 
instructed to stop work immediately 
during the power interruption. An- 
other example is the requirement that 
the exhaust and supply fans for P4 fa- 
cilities be interlocked. This assures 
that in the event of failure of the ex- 
haust fan, the supply fan will auto- 
matically shut down, preventing the 
pressurization of the laboratory envi- 
ronment. As with the class III cabinet 
example, personnel would stop their 
work because of the loss of secondary 
protection provided by the ventilation 
systems. Operational procedures, 
therefore, become an important ele- 
ment in assuring safety in the event of 
any system failure. 
Institutions are required to devise 
emergency plans to handle possible 
problems. In response to recommenda- 
tions of the Environmental Protection 
Agency Study Group on Recombinant 
DNA and to concerns raised by com- 
mentators, NIH has stipulated more 
clearly (in the supplement to the 
PRG-NIH entitled, “Laboratory 
Safety Monograph”) certain elements 
in these emergency plans. Moreover, 
NIH staff have recently met with rep- 
resentatives of the center for Disease 
Control (CDC) to establish a mecha- 
nism for providing advice, consulta- 
tion, or assistance, if necessary, in case 
of an emergency, such as an accident 
in the laboratory. 
Laboratory sajety. A number of com- 
mentators felt that the PRG-RAC was 
vague in regard to the training in 
safety of researchers, students, and 
janitors. It was urged that specific cur- 
ricula be developed and that a require- 
ment for certification of training be 
stipulated in the guidelines (a recom- 
mendation also made by the EPA 
Study Group on Recombinant DNA). 
It has been suggested, further, that 
NIH develop curricula for training. 
At the present time. NIH has a con- 
tract with the American Society for 
Microbiology (ASM) to develop mini- 
mum standards for training partici- 
pants in recombinant DNA research. 
The ASM Working Panel will consider 
what standards of training in micro- 
biologic techniques are appropriate for 
the conduct of experiment requiring 
PI through P3 containment condi- 
tions. The Panel will solicit views from 
the scientific community to develop 
minimum requirements for training. 
The Panel’s report will be made availa- 
ble to the IBC’s and investigators to 
set standards for all who participate in 
this research. In view of these develop- 
ments, formal certification require- 
ments by NIH are considered prema- 
ture. 
Other commentators stressed the 
need for more stringent measures in 
regard to safe practices. In particular, 
these commentators urged regular 
monitoring of laboratory facilities, 
preferably at all P levels. This would 
include monitoring of microbiological 
practices, serological monitoring, and 
CDC review of incidence of infections. 
It was also suggested that regular in- 
spections be performed by individuals 
not associated with the institution (to 
preclude conflict of interest); that the 
guidelines require a member of the 
work force to be represented on the in- 
stitutional biosafety committee; and 
that penalties (other than cutoff of 
funds) be imposed on violators as a de- 
terrent. I have accepted many of these 
proposals; the specific NIH actions in 
regard to them are discussed in Part 
IV of this document. 
Appendix D, "Supplementary Infor- 
mation on Physical Containment,” was 
added to the 1976 guidelines in re- 
sponse to numerous requests for great- 
er specificity in describing contain- 
ment requirements. Commentators 
noted the absence of this document 
from the PRG-RAC and urged that it 
be retained and further expanded. Ac- 
FEDERAL REGISTER, VOL 43, NO. 146 — FRIDAY, JULY 28, 1978 
[ 14 ] 
