The Associated Press is reporting that many nations, in particular the USA, have changed their surveillance methods for keeping track of Swine Flue (H1N1), and are no longer counting confirmed cases. The justification for this is that the confirmed cases count was already massively underestimating the numbers affected, and in any case, it is no longer useful once the disease hits a certain proportion of the population. This may be true on a whole population level, but the move away from counting cases means that changes in particular populations and areas below subnational level are less observable – and this is a problem if the disease is affecting some groups and places more than others. It might for example be crucial to deciding who and where receives vaccinations, for example. There is also the added complication of budget cuts in local government surveillance resulting from the recession. As with many kinds of caring surveillance, one key question is not whether the surveillance is perfectly accurate, but whether the surveillance is ‘good enough’ for the purpose for which it is intended, and in the case of diseases, this is sometime a tricky thing to determine.
Internet disease tracking using interactive maps or mash-ups seems to be be one of the more constructive uses of the surveillance potential that comes with the combination of easy-to-use digital mapping and online communications. Both Computer World and The Guardian tech blog reported a few days back how Google, following on from its use to track previous flu epidemics, is experimenting with tracking swine flue cases in Mexico.
However other web-crawler-based systems also exist for tracking the spread of disease (or indeed potentially almost anything) as The Guardian reported on Wednesday. Leading the way is HealthMap, which comes complete with Twitter feeds and suchlike.
As the latter report makes it clear however, this is not all just good news; there are many problems with the use of web-crawlers in providing ‘reliable’ data not least because the signal to noise ratio on the Internet is so high. The other problem is that although the might appear current or even ‘predictive’ by virtue of their speed and interactivity, they are of course actually always already in the past, as they are compilations of reports many of which may already be dated before they are uploaded to the ‘net. Better real-time reporting from individuals may be possible with mobile reports, but these could lack the filter of expert medical knowledge and may lead to the further degredation in the reliability of the data. Can you have both more reliability and speed / predictability with systems like this? That’s the big question…
(Thanks to Seda Gurses for pointing out the CW article to me!)
Without a doubt, there has been a massive jump in the appearance of surveillance in the news worldwide this week. However this may have escaped the notice of many people who are interested in ‘surveillance’ as social scientists would conventionally understand it. The surveillance in question has been that around the current global spread of the H1N1 virus variant, and the Swine Flu disease which it causes. Perhaps the main emerging question as the H1N1 spreads around the world, aided by our increasingly interconnected global transport network and the mobility for multiple reasons, of both people and animals, is whether these surveillance systems have been adequate. This question is addressed by the New York Times today, for example.
The strange thing for me, as I once again discovered this week, is that many surveillance studies people, who are used to looking at everything from CCTV to Internet monitoring, don’t think of this, or aren’t used to thinking of this as surveillance. The argument was made to me this week that one shouldn’t be fooled by the coincidence of words, that medical / epidemiological surveillance is something quite different from the ‘systematic, focused attention to personal data for the purposes of influencing behaviour’ (or whatever definition happens to be used). Partly this is because these definitions are inadequate in the first place. Surveillance – as Dr Andrew Donaldson of the Centre for Rural Economy at Newcastle, who specialises in looking at the geography of the nonhuman, and I wrote back in 2004 in an article called ‘Surveilling Strange Materialities’ in Society & Space – is primarily a mode of social ordering. This involves relationships of humans and things. ‘Things’ are always involved: whether they are technological things that originate within society, or natural things that originate without but enter or are are brought in to a social relationship. When these things are regarded as threatening, as with viruses or other kinds of pathogens, they are subjected to the kinds of activities that would be quite familiar to surveillance studies academics: there is monitoring, collection of data, analysis and matching, and action that results in real ordering effects, that takes through global, regional and national networks of scientists and research centres. This is surveillance and the ways in which medical science and social science understands it are not separate concepts that happen to share the same word.
The current regime is the result of the experience of previous outbreaks, and there is now increasing connection between the systems set up by the OIE, or the World Animal Health Organisation as it is now increasingly called, and the World Health Organisation with which most of us are more familiar. A lot of this simply involves keeping a close watch on the patterns of reported diseases and illness and comparing these with the normal levels of disease. In the case of swine flu, unusual patterns were reported in Mexico, but by the time this had been confirmed as something of real concern, people from the affected areas were already moving around the world, coming into contact with other people and spreading the virus on to new hosts. This is hardly the fault of the surveillance systems in operation. Disease surveillance is a far more complex operation than monitoring a city-centre with a CCTV camera. It relies first of all on disease outbreaks being recognised as disease in the first place – not always easy when many originate in places with so many species of human misery and suffering. Then, the disease much be identified and some assessment must be made of the risks it poses. One interesting thing about viruses is that the viruses that spread most rapidly tend to be those that do not prove fatal to large numbers of people (or at least not quickly). The reason for this is clear. Viruses are like little packets of swiftly mutating genes and evolutionary ‘strategies’ come and go very quickly. But they have very limited ability to move and reproduce without a medium of some kind, in other words a host. If in their reproduction they kill the host very quickly, their subsequent movement and further reproduction could be very limited. This is one of the reasons why fearsome diseases like Ebola have thus far stayed fairly localised. So, from a human or animal perspective, rapidly-spreading viruses may not be of much extra concern (as with the ordinary waves of flu that pass through northern countries in winter). However, because viruses evolve by mutation and swapping genes between themselves and other similar viruses, there is always the possibility of a new strain of any virus that is both highly virulent and ultimately fatal.
The question is then, how much to panic when a new or unusual outbreak is spotted. However because we are talking both about the chances of this being spotted, reported, the reports analysed by regional centres, and then probably confirmed through more systematic sampling, this time-frame almost inevitably cannot be made to coincide with the rapid and global mobility of human and animal populations – and with animals I am not just talking about the wild bird movements that so concern disease specialists in the case of H5N1 (Bird Flu), but the global trade in animals for all sorts of purposes from food to pets.
The other interesting issue is the indirectness of the surveillance involved. We’re not generally tracking the virus itself (except at the laboratory level), but the carriers, the animals or humans infected with the virus. It’s the only thing we can do on the whole. You can’t monitor viruses in the world. Humans and large animals are much easier, and since it is the effects of these viruses on these organisms that we are concerned about in the case of H1N1, much more relevant. The next point then is the altering of behaviour. The control of the spread of large outbreaks as they threaten to turn into pandemics, must involve slowing down or stopping the movement of people and animals that facilitates the spread of diseases. This means the kind of authoritarian controls, and indeed disruptions to flows of people and goods, that we would regard as utterly unacceptable in most other situations because we accept the expert judgment of the seriousness of the risks…