 Membership
 Publications
 Meetings
 Competitions
 Community
 Programs
 Students
 High School Teachers
 Faculty and Departments
 Underrepresented Groups
 MAA Awards
 MAA Grants
 News
 About MAA
In Part 5 we took it for granted that the parameters b and k could be estimated somehow, and therefore it would be possible to generate numerical solutions of the differential equations. In fact, as we have seen, the fraction k of infecteds recovering in a given day can be estimated from observation of infected individuals. Specifically, k is roughly the reciprocal of the number of days an individual is sick enough to infect others. For many contagious diseases, the infectious time is approximately the same for most infecteds and is known by observation.
There is no direct way to observe b, but there is an indirect way. Consider the ratio of b to k:
b/k  = 
b x 1/k 
= 
the number of close contacts per day per infected 

= 
the number of close contacts per infected individual. 
We call this ratio the contact number, and we write c = b/k. The contact number c is a combined characteristic of the population and of the disease. In similar populations, it measures the relative contagiousness of the disease, because it tells us indirectly how many of the contacts are close enough to actually spread the disease. We now use calculus to show that c can be estimated after the epidemic has run its course. Then b can be calculated as c k.
Here again are our differential equations for s and i:
We observe about these two equations that the most complicated term in both would cancel and leave something simpler if we were to divide the second equation by the first  provided we can figure out what it means to divide the derivatives on the left.
The differential equation in step 1 determines (except for dependence on an initial condition) the infected fraction i as a function of the susceptible fraction s. We will use solutions of this differential equation for two special initial conditions to describe a method for determining the contact number.
Three features of this new differential equation are particularly worth noting:
There are two times when we know (or can estimate) the values of i and s  at t = 0 and t = infinity. For a disease such as the Hong Kong flu, i(0) is approximately 0 and s(0) is approximately 1. A long time after the onset of the epidemic, we have i(infinity) approximately 0 again, and s(infinity) has settled to its steady state value. If there has been good reporting of the numbers who have contracted the disease, then the steady state is observable as the fraction of the population that did not get the disease.
David Smith and Lang Moore, "The SIR Model for Spread of Disease  The Contact Number," Loci (December 2004)
Journal of Online Mathematics and its Applications