Statistics from Altmetric.com
THE VETERINARY RECORD is published by the British Veterinary Association. It was founded in 1888 by William Hunting, FRCVS.
EFFECTIVE surveillance is fundamental to disease control. This is true for avian influenza as for other diseases, where early detection and rapid implementation of control measures are crucial if the spread of disease is to be limited. If you're searching for something and know what you're trying to find (which is not always the case in disease surveillance), it is useful to know where best to look, which is where targeted surveillance can help. Targeting surveillance can increase the chances of detecting disease, and help make best use of resources. In a paper on pp 775 to 781 of this issue, a research team made up of epidemiologists and ornithologists describe how they combined available data on wild birds and poultry holdings in Great Britain to develop a risk-based approach to surveillance for h5n1 avian influenza in wild birds. The work was undertaken in 2006 in response to increased concern about the presence of h5n1 avian influenza in Europe and the potential risks to the uk poultry industry. Its aim was to identify the geographical areas where surveillance, notably for the collection of dead birds for screening, should be targeted. The results are already finding practical application in the form of a map on the avian influenza section of defra's website (www.defra.gov.uk/avianflu), identifying counties where surveillance for avian influenza in wild birds is considered a priority.
Making use of information held on the Great Britain Poultry Register, together with data from various ornithological monitoring schemes on the distribution and abundance of 24 wild bird species considered most likely to introduce the virus into Great Britain, the researchers set out to develop a risk profile to identify the areas where h5n1 is most likely to enter and spread to commercial poultry. As in many surveillance studies, the data available were not complete and this, coupled with gaps in the knowledge of the epidemiology of avian influenza, meant that various assumptions had to be made. Nevertheless, by assigning risk scores to different types of poultry holding, as well as abundance scores to each of the 24 wild bird species during each month of the peak migration period between October and December, the researchers were able to construct a priority map for surveillance for each 10 km square of Great Britain. Squares were ranked on a scale of 1 to 6, with a rank of 6 indicating that there were either no poultry or wild birds present in the 10 km square, and the ranks from 5 to 1 indicating risks in increasing order of priority for surveillance. Their results indicated that surveillance would be best focused on areas of Norfolk, Suffolk, Lancashire, Lincolnshire, south-west England and the Welsh borders. These areas have significant poultry populations, including a large number of free-range flocks, and a higher abundance of the 24 wild bird species. Areas of lower priority included Anglesey, southwest Wales, north-east Aberdeenshire and the Firth of Forth area of Scotland, which have significant numbers of the wild birds and dense but localised poultry populations.
The authors point out that their maps should be regarded as adaptive, in that the thresholds at which the squares are ranked as high priority may be reviewed, depending on the number of birds reported and the resources available for collecting and analysing them. The maps can also be updated to incorporate new wild bird or poultry data, or additional species if the international disease situation changes or new information becomes available. The researchers emphasise, however, that because of the qualitative nature of the work, and the many uncertainties affecting the risk of incursion, the maps cannot be used to predict where the next outbreak might occur or where the next infected bird might be found. They warn that poultry producers, wherever they are located, must maintain biosecurity and good management practices, to minimise contact between their poultry and wild birds, and thus reduce the risks of an outbreak.
The work described in the paper predates last month's outbreak of h5n1 avian influenza in poultry in Suffolk, epidemiological investigations into which are continuing. A preliminary epidemiology report published by defra last week was unable to pinpoint the source of the outbreak. It noted, however, that investigations to date had provided no evidence that infection was introduced via imported poultry or poultry products, or any activities associated with such imports. No h5n1 infection had been detected in wild birds in the area, nor were there any incidents of high mortality among wild birds observed. However, the report notes that the first infected premises were located in an area where wild birds were relatively common and close to an ornamental lake supporting about 1000 waterfowl. Wild birds, most likely migratory species from central Europe, could not be ruled out as a source of infection, and an enhanced surveillance programme had been initiated (see p 766 of this issue).
Targeted surveillance is likely to play an increasingly important role in disease control in the future, as animal heath strategies are refined and epidemiological techniques become more sophisticated. At the same time, it is essential that background ‘scanning surveillance’ is maintained, to spot new diseases that may not be expected. Forewarned is forearmed and, like previous papers in this journal anticipating the arrival of bluetongue, the work described in this issue shows how research can contribute practically to efforts to prevent and control disease. Meanwhile, experience of disease outbreaks continues to demonstrate the importance of epidemiological studies, before, during and after the event.