Epidemic threshold (source code)

= Epidemic threshold

An <epidemic threshold> separates conditions under which a rare infection initially declines from conditions under which it can grow. It is usually obtained by studying the <linear stability> of the <disease-free equilibrium>, holding the susceptible background at its disease-free value. The <basic reproduction number> organizes this calculation: under the usual <next-generation matrix> assumptions, invasion is possible when its value exceeds one.

For a two-group model with only cross-group transmission, suppose the small infected abundances satisfy
$$
\dot i_1=-\gamma_1 i_1+b i_2,\qquad
\dot i_2=c i_1-\gamma_2 i_2,
$$
with $b,c,\gamma_1,\gamma_2>0$. The <Jacobian matrix> has negative <trace> and <determinant> $\gamma_1\gamma_2-bc$. Thus one <eigenvalue> becomes positive when $bc>\gamma_1\gamma_2$, while both have negative real parts when the inequality is reversed. Equivalently, this model has $\mathcal R_0=\sqrt{bc/(\gamma_1\gamma_2)}$ and threshold $\mathcal R_0=1$. This is a local invasion test; the ultimate outbreak size and nonlinear equilibria require the full model.