- 9 - 
Lapp^ 
or that pipettes which are stood upright in sterilizing medium 
or blown out to clear the last drop are high-risk devices for 
creating Infectious aerosols. Nor will they necessarily recognize 
that snapping the neck of an ampoule of stored biological material, 
or merely lifting the lid of a petri dish will generate potentially 
infectious aerosols of the agents they contain. 
These and a host of as yet unrecognized sources of human error 
and hazard have undoubtedly contributed to the contamination of an 
Incredibly high percentage of all the cell cultures in the world 
with one single line of fibroblast cells known as HeLa. The bottom 
line of any risk estimate must certainly take into account that 80% 
of all reported laboratory accidents have no known cause. Neither 
the NIH guidelines nor the environmental impact statement satis- 
factorily acknowledge the element of human error or incorporate it 
into their risk equations. 
So where are we? Somehow it all sounds familiar. We are 
faced with a high technology which promises unforeseeable benefits. 
Ostensibly, to reap them, all that need be done is to ensure that 
the immediate hazards of its use be regulated to the satisfaction 
of the scientific community. Yet the large-scale, second-order 
consequences of mishaps are ill-considered. Once an error has been 
made, cessation of a recombinant DNA experiment in no sense ensures 
that the diffusion of the hazardous agent will stop. Unlike any 
other risk estimation techniques we have used in the past, recombinant 
Appendix K — 18 
