"Christmas has come early for disease modellers. Biomodels can predict the future of epidemics because they don't pull apart the biology to substitute it with maths."

The predictive capabilities of biomodels stem from regularly monitoring wide-reaching disease factors in a population, such as subclinical disease. For example, most infectious diseases have a subclinical form that initially doesn't show clinical signs, and when the signs are exhibited they are usually mild. There is also an acute form of disease, which is exhibited relatively quickly after initial infection. Hence the acute form can be controlled where the disease is diagnosed early, and insufficient time has elapsed for the disease to spread to other indivduals or groups of individuals. However, it is the subclinical form which spreads disease without showing cinical signs, and seeds an epidemic (at the regional level) or seeds an outbreak (at the local level). Hence, by measuring the initial level of subclinical disease, it's possible to predict disease prevalence (or the final number of infected individuals): this is because prevalence has already been measured (for past epidemics) at different levels of subclinical disease. There is a high degree of accuracy in predicting disease prevalence from subclincial disease [15]. Epidemiologists have identified similar biomodels in the literature ie. [22],[23], etc.

The same principles apply to predicting the duration of an epidemic (or outbreak).

Subclinical disease: