The mean infective period in a SIR Model

146 Views Asked by At

I'm going to quote the relevant passage in the textbook and then ask my question.

The assumption that the infectives leave the infective class at rate $\alpha I$ per unit time requires a fuller mathematical explanation, since the assumption of a recovery rate proprotional to the number of infectives has no clear epidemiological meaning. We consider the "cohort" of members who were all infected at one time and let $u(s)$ denote the number of those who are still infective $s$ time units after having been infected. If a fraction $\alpha$ of these leave the infective class in unit time, then

$$ u'=-\alpha u$$

and the solution of this elementary differential equation is

$$u(s)=u(0)e^{-\alpha s} .$$

Thus, the fraction of infectives remaining infective $s$ units are having become infective is $e^{-\alpha s}$, so that the lenth of the infective period is distributed exponentially with mean $${\color{red}{\int^{\infty}_{0} e^{- \alpha s} \ ds = \frac{1}{\alpha}}}.$$

Now if the fraction of infectives remaining infective s units are having become infective is $e^{-\alpha s}$, then the fraction no longer infective $s$ units after having become infective is $1-e^{-\alpha s}$ which is the cdf of an exponential distribution.

Then since the pdf of an exponential distribution is $\alpha e^{- \alpha s}$, isn't then the mean infective period $$E(S) = \alpha \int_{0}^{\infty} se^{-\alpha s} \ ds = \frac{1}{\alpha} \ ?$$

1

There are 1 best solutions below

7
On BEST ANSWER

For a non-negative random variable with density function $f(x)$ and cumulative distribution function $F(x)$, one way to calculate the mean is $\displaystyle \int_0^\infty x f(x) \; dx$ but another is $\displaystyle \int_0^\infty (1-F(x)) \; dx$.

You used the former while the book used the latter to get to the same result.

Note that you could write $E[X]= \displaystyle \int_0^\infty S(x) \; dx$ where $S(x)=1-F(x)$ is the survival function.