Google recently has been used to track the development of the flu by using the number of queries about flu symptoms and treatment as a measure of the spread of the disease. However, this is what is known as a lagging indicator because it reflects what has already happened, not what will happen in terms of propagation of the disease (although some argue that the Google measure is a contemporaneous measure of disease). The CDC methods have a tendency to lag the outbreak by 1-2 weeks.
A different approach was tried recently by researchers to ascertain if it might help to predict the spread of a disease. The strategy sought to make use of what is known as the "friendship paradox" -- the friends of any random individual are more likely to be central to the "social web" (grouping of friends and acquaintances) than the individual himself/herself. The approach was to identify these central figures, those that are "more connected" to groupings of others, and therefore are more likely to catch diseases such as the flu early. The theory was that this would allow health authorities to spot outbreaks weeks in advance of current surveillance methods.
Using this method researchers monitored the spread of both the seasonal flu and the "swine flu" (H1N1) through students and their friends at Harvard. They found that infection rate peaked two weeks earlier among the group of "more connected" individuals; their social links were thus apparently causing them to become infected earlier.
Since this was developed with the advantage of hindsight, the researchers went back and compared diagnoses between the "more connected" and the "random" groupings. A statistically significant difference was first detectable 46 days before visits to the health service peaked for the random group. For those with self-reported symptoms, the difference was 83 days.
This is at best a proof of potential concept, given the limited number of individuals in the study and its testing in the social milieu of a university. However, it may have promise as a method.
A good report on the study can be found at http://www.sciencedaily.com/releases/2010/03/100308151049.htm.