A cold. Another one. A fever. The neighbour has it too. Pain in the limbs. The entire neighborhood is laid low. The flue. Every year. Diseases as calculable as the seasonal flue and their epidemiological data for disease incidence are interesting for physicians treating those diseases. The installed flue surveillance systems such as the Sentinel-System or the RealFlu-System have exactly that purpose. But in case of a serious threat we would appreciate to hear about it a bit earlier. If in a house or a street or any institution typical symptoms heap, the alert bells might go off, even prior to a final diagnosis.
The pandemic plans of the various countries account for this matter in so far as such facilities as kindergartens are constrained to keep an eye on heaping symptoms respectively to close the place if need be. The website Whoissick.org, published recently in the internet, has a similar approach. The concept is brilliantly obvious. Whoever is sick, can get registered on the site and enter his symptoms. By aid of the available functions of Google Maps, an entry is generated on a city map respectively a topographic map, exactly in position, where the sick person is located. Now every visitor of the site can take a look and – in the ideal case – see, which medical problem occurs where. The information provided can be analyzed with a remarkably easy-to-use statistics tool. For instance, it is possible to select any map excerpt and find out, how many persons had which symptoms over the last few weeks in order to draw conclusions on any heaping occurrence of certain diseases.
A bustle ground for hypochondriacs
So much for theory. Whoissick.org has generated enormous clicks with this concept during a few weeks, mainly in the US. In other countries, this site has only a few visitors. If you take a look at the individual – of course anonymous – patient, you will find people seriously interested in informing about their medical conditions. But you will also find a significant number of obvious hypochondriacs, wisecrackers and drunkards. In Berlin, for example, you find a chatty fellow citizen reporting in several posts, that he drank too much at the barbeque and now feels “totally sick”. Entries like this have fuelled a wide debate about sense and nonsense of Whoissick.org, especially in some English speaking countries. The reproach, thrown into the discussion by the co-ordinator of the Australian influenza surveillance program, sums down to the following: The data is worthless if not collected professionally, such opening floodgates for manipulation. Statistics based on such dubious data would end up being pseudo-scientific. Other objections target on an epidemiology based more on subjective symptoms than on definite diagnoses or at least objective medical evidence, which does not have enough selectivity to allow any sensible statement at all.
Algorithms against data garbage?
A certain arrogance of the experts is shining through all those arguments. And according to expectations, this has brought several vehement defenders of the website on the scene, mainly bustling about in various weblogs. One look at the blog-search engines Technorati or the Google blog-search attest the considerable demand for discussion. Only little argues against keeping things in perspective and trying to get to the bottom of the potential of this tool, for example during the next flu-season. Whoever wants an epidemiology based on the people self-reporting, will have to live with hypochondriacs and wisecrackers. But it could be worth a shot to develop mathematic algorithms to filter this type of entries like background noise and to simply test the resulting tool during any epidemic in real life. If a local break-out can be discovered only one or two days earlier, it might be worth it….