National Institutes of Health 
Dr. Fredrickson 
9/21/78 
Page 4 
of the height of the building and an exhaust velocity upwards 
of at least 2,500 fpm. 
2. There shall be no taller buildings, hospitals, or 
densely populated areas within a \ mile radius. 
As a practical matter, the above recommendations are 
tantamount to a recommendation that the exhaust air be HEPA 
filtered in almost all cases. The question will undoubtedly 
be asked, "Is this necessary?". 
Experience with clinical microbiology laboratories in 
hospitals, most of which do not approach the P3 guidelines, 
suggests that the spread of infection from laboratory ventilation 
systems is nil. On the other hand, these laboratories generally 
deal with class 3 agents that are not highly contagious . (do not 
cause epidemics) such as M. Tuberculosis, and whose etiology 
is well known. The purpose of P3 containment as we understand 
it, is to reduce the probability of escape and replication of 
a viable organism containing recombinant DNA to a very small 
number. Levels of escape which are acceptable for organisms 
such as M. Tuberculosis are not acceptable for organisms from 
a P3 laboratory. We therefore conclude that the above recommend- 
ations are indeed necessary. 
E. Section III-B-5 of the Laboratory Safety Monograph 
specifies a leak test procedure for positive pressure plenums in 
Class II Type A cabinets. We have two comments relating to this 
paragraph 
1. We understand that NIH does not use this procedure for 
Class II cabinets in Bethesda, but uses instead a pressure test 
specified by Test Protocol No. 16 to RFP 263-76P (62' -0125cc. We 
recommend that this protocol be listed as an alternate procedure. 
2. We have found that pressure plenums on Class II cabinets 
can be tested using DOP. The procedure is similar to paragraph 
III-B-7 of the Monograph except that gaskets and joints outside 
the cabinet are scanned. Without turning on the DOP, these areas 
are scanned to determine a background reading. Then the DOP is 
turned on and the scanning repeated. An increase reading of 
.01% or more indicates a leak. The procedure is applicable in 
any laboratory where the background reading does not exceed .1% 
(less than 1% of the laboratories we test have reading above 
this value) . A linear photometer is required. This procedure 
is simpler than the Halogen test, eliminates the release of 
fluorocarbons into the atmosphere, and is consistent with the 
test requirements for the exhaust HEPA filter. We suggest that 
this procedure be added to the Laboratory Safety Monograph. 
[A-221] 
