Biomedical Engineering Reference
In-Depth Information
Figure 4
.
Social quarantine
. To model avoidance of sick individuals by the healthy, uninfected
individuals reduce the probability of realizing a contact with sick individuals in (
B
). However, the
probability reduction must be substantial to protect a population from extinction (
C
). Social with-
drawal by infected individuals (
D
) is even more effective in containing an epidemic because they
often detect illness before signs are apparent to others (e.g., after 1 time unit of infection, vs. 2 in
B
).
viduals are killed. However, uninfected individuals must detect and avoid those
who are infectious with an extremely high rate of success to halt an epidemic.
Figure 4C shows that under the conditions of panels A and B, uninfected indi-
viduals must detect and evade at least 80% of infected individuals to avoid con-
sistent host population extinction.
Surprisingly, the network dynamics that most decisively contain disease do
not stem from the self-protective behavior of the healthy, but from the involun-
tary behavior of the sick. It has recently been discovered that proinflammatory
cytokines—the signaling molecules that initiate an immune response—also
prompt the brain to unleash an integrated package of "sickness behaviors" that
immobilize us with fatigue, malaise, and myalgia, and substantially crimp our
social and reproductive motivation (2,3). Sickness behavior has generally been
analyzed in terms of its advantage for individual recovery, but its most signifi-
cant contribution may lie in protection of the group. Even small reductions in
contact can substantially impede the spread of infection through a sparse net-
work. Figure 5B shows an example in which infected individuals withdraw
10% of their social contacts at random, resulting in substantial increases in
population survival relative to a comparable scenario in which the sick do not
withdraw (5A).
Self-generated quarantines are more efficient than socially imposed ones
because we generally feel sick sooner than we look it, and this allows a faster
blockade of transmission links. Figure 4D emphasizes this point by comparing
the effects of a fixed probabilistic reduction in contact generated by the sick
individual one time unit following infection with the same probabilistic reduc-
tion generated by healthy individuals once they recognize illness at two time
units post infection (4B). A sick host's self-generated quarantines are also more
efficient than social quarantines because the motivation for altered behavior
