by Angela Herring

In 2011, sci­en­tists suc­cess­fully engineered a lethal avian flu virus to be trans­mis­sible between birds as well as mam­mals and pos­sibly humans. The novel virus, a genet­i­cally engi­neered vari­a­tion of H5N1 avian influenza, sparked an enor­mous debate among both the research com­mu­nity and the public about how to manage such research and whether it should even be car­ried out at all.

That’s where North­eastern stepped in. “We thought it was impor­tant to pro­vide some hard num­bers to the debate,” said Alessandro Vespig­nani, a world-​​renowned sta­tis­tical physi­cist and the Stern­berg Family Dis­tin­guished Uni­ver­sity Professor.

In a paper released Thursday in the journal BMC Med­i­cine, Vespig­nani and his col­lab­o­ra­tors pro­vide those hard numbers—and they aren’t ter­ribly reas­suring. “This study pro­vides a very accu­rate mod­eling approach to assess the prob­a­bility of con­tain­ment in the case of acci­dental escape,” explained Vespig­nani, who holds joint appoint­ments in the Col­lege of Sci­enceCol­lege of Com­puter and Infor­ma­tion Sci­ence, and Bouvé Col­lege of Health Sci­ences. “Unfor­tu­nately there are large chances that the out­break will not be contained.”

Vespig­nani and his research team used census data from the city of Rot­terdam in the Nether­lands to create a com­pu­ta­tional model that tracked how an exper­i­mental virus would spread if it were acci­den­tally released from a facility oper­ating at a biosafety level of 3 or 4. These labs carry out the most health haz­ardous bio­log­ical research in the world and are often located in pop­u­lous urban areas.

The team, which includes col­lab­o­ra­tors at the Bruno Kessler Foun­da­tion in Trento, Italy, looked at the effec­tive­ness of sev­eral non-​​pharmaceutical inter­ven­tions fol­lowing a release: iso­la­tion of the lab­o­ra­tory; house­hold quar­an­tine of the infected worker; and quar­an­tine of indi­vid­uals who came into con­tact with the infected worker.

They exam­ined these con­tain­ment sce­narios in the con­text of dif­ferent viruses, ranging in trans­mis­si­bility from a mild sea­sonal influenza to one sim­ilar to the Spanish flu, which killed nearly 5 per­cent of the world’s pop­u­la­tion in 1918.

The results of the sim­u­la­tion sug­gest a 5–15 per­cent chance that an acci­dental escape would not be detected, espe­cially in the case of very trans­mis­sible viruses and those where symp­toms are not imme­di­ately spotted. In addi­tion, they found that con­tain­ment would depend on the struc­ture and den­sity of the local pop­u­la­tion sur­rounding a facility.

“Most BSL labs are in big urban areas,” Vespig­nani explained. “In those areas we show that the prob­a­bility of not con­taining the out­break is three to five times larger than what it would be in iso­lated areas.”

While the prob­a­bility of acci­dental release is extremely low—there’s only 0.3 per­cent chance of a virus escaping one of these labs each year—even a single event can trans­late into a vast public health emer­gency, said Ste­fano Merler, one of the researchers who is based at the Kessler Foun­da­tion. More­over, the number of BSL3 and 4 lab­o­ra­to­ries is increasing, cre­ating a greater com­bined risk the world over.

As such, this type of research could prove extremely useful for pol­i­cy­makers deciding where to build facil­i­ties, Vespig­nani said. “It’s also useful in weighting the oppor­tu­nity of autho­rizing exper­i­ments and in defining pro­to­cols for con­tain­ment,” he added.

Originally published in news@Northeastern on November 28, 2013.