Epidemic modeling: What you need to know

Epidmic models have been around for ages and have been imperative to learning alot of what we today hold to be obvious. The data from real outbreaks are sometimes not good enough and scientific ethics restrict us from conducting live experiments - on good grounds.
One prime example of the use of epidemic models is the realization that epidemics critical phenomenons that can only occur once the conditions are "right". The infective disease has to be contagious enough, people have to get together frequently enough and they have to stay infectious long enough to infect their neighbors. Once the conditions are met, the critical value, the famous R0, will climb above 1.0 and the epidemic will take of like a nuclear chain reaction. The critical value can in simple cases be calculated by use of simple mathematical epidemic models. In more complicated scenarios we use computers.

The use of models, either just pen-and-paper models or computer models is common of all sciences. What make epidemic models so special is that they depend completly on what we know about sociology, how people behave. People behave unpredictably. They have random contacts with each other. Even if we could get down exactly what everybody was doing on a particular day, there's no telling that that's what they will do on any other day. There could be a big game on TV. Or why not a major disease outbreak prompting everybody to stay at home out of fear of getting infected. As Murray Gell-Mann put it, sociology is as complex as physics would be "if particles could think".

There's definitely no logic,
To human behaviour.

Björk

Next page