When people take pre­cau­tionary mea­sures, their behavior often changes as a result,” said Ravi Sun­daram, an asso­ciate pro­fessor in the Col­lege of Com­puter and Infor­ma­tion Sci­ence at North­eastern Uni­ver­sity. For example, people wearing seat­belts may drive faster, people who’ve received flu shots may skip washing their hands before eating, and those who’ve been vac­ci­nated against sex­u­ally trans­mitted dis­eases may tend to engage in more risky behavior.

But how do these behav­ioral changes affect a contagion’s spread when the pre­cau­tion is not fool­proof? This is pre­cisely the ques­tion that Sun­daram and his col­league, pro­fessor Raj­mohan Rajaraman asked in research recently pub­lished in the Journal PLOS ONE.

The duo used net­work sci­ence tech­niques to look specif­i­cally at the influenza vac­cine, which has a 20 to 40 per­cent chance of failing, and the HIV/​AIDS anti­retro­viral, which is unsuc­cessful 25 to 75 per­cent of the time.

Given these sta­tis­tics, Rajaraman said, “if the level of risky behavior exceeds a cer­tain threshold, then you arrive at some strange sce­narios in which the more you inter­vene the worse you’re making it.”

The flu virus only requires that one person make a poor deci­sion before it jumps to another host. In this case, if the vac­ci­na­tion rates are low, then increased vac­ci­na­tion will tend to make the problem worse. This anomaly does not occur, how­ever, when a large frac­tion of the pop­u­la­tion is vac­ci­nated, which will out­weigh the neg­a­tive impacts of risky behavior.

How­ever, things aren’t this simple with STDs, which require two par­ties’ par­tic­i­pa­tion to be trans­mitted from one host to another. Here, risky behavior is over­come by increased vaccinations—but only to a point. Using a com­pu­ta­tional model based on data from the New River Valley in Vir­ginia, the researchers found that after about 40 per­cent of the pop­u­la­tion was vac­ci­nated, the tables turned. Now the risky behavior out­weighed the pos­i­tive impacts of the vaccine.

What hap­pens in the HIV case is that since the risky behavior requires con­sent on both sides, the number of risky inter­ac­tions for low levels of vac­ci­na­tion is still very small,” said Rajaraman. “So you ben­efit with increased vac­ci­na­tion. But this may no longer hold at high levels of vac­ci­na­tion.” That is, one person’s risky inten­tion will not trans­late into risky behavior until the second party is also vaccinated.

The crit­ical take­away of the research, Sun­daram said, “is that we have to have some kind of behav­ioral inter­ven­tion that is coor­di­nated with the med­ical inter­ven­tion. So we tell people go get a flu shot, but then you also need to accom­pany it with some kind of behav­ioral inter­ven­tion in which you tell people to be aware that it isn’t perfect.”

The research is part of an ongoing col­lab­o­ra­tive study with com­puter sci­en­tists at Vir­ginia Tech and epi­demi­ol­o­gists at Penn­syl­vania State Uni­ver­sity in which the team is looking at sim­ilar prob­lems across a variety of contagions.

The results, if con­firmed in real-​​world set­tings, could have sig­nif­i­cant policy impli­ca­tions, said epi­demi­ol­o­gist Stephen Eubank, one of the team’s col­lab­o­ra­tors at Vir­ginia Tech University.