• Report
• Appendix 1 Mandate from the European Commission
• Appendix 2 Nature and severity of disease in sheep and cattle 
• Appendix 3 Within herd distribution of infection
• Appendix 4 Time space distribution of virus spread between herds
• Appendix 5 Role of environmental factors – wind analysis
• Appendix 6 Role of environmental factors – small scale environmental analysis
• Appendix 7 Role of environmental factors – large scale environmental analysis
• Appendix 8 Role of human intervention
• Appendix 9 Distribution and dynamics of vector species

Executive Summary

Bluetongue (BT) is an arthropod-borne viral disease of domestic and wild ruminants, affecting particularly certain breeds of sheep with severe clinical disease, including mortality. On 14 August 2006, a private veterinary practitioner in the southern province of Limburg, in The Netherlands, notified the veterinary authorities of BT-suspect cases on four different holdings in that Member State (MS) These were the first suspicions of a rapidly-spreading BT virus (BTV)-epidemic in north-western Europe, which has since affected cattle and sheep holdings in Belgium, Germany, France, Luxembourg, and The Netherlands. On 28 August 2006, the CRL in Pirbright announced that BTV-serotype 8 (BTV-8) was causing the outbreaks.

The veterinary authorities of the three initially-affected countries (Belgium, Germany, and the Netherlands) decided at the onset of the BTV-epidemic in north-western Europe to form a research group with epidemiologists from these countries in order to provide science-based decision support for future BTV monitoring and surveillance. The European Food Safety Authority (EFSA) was requested by the European Commission (EC) to carry out a global epidemiological analysis of the ongoing outbreak. Therefore a BTV-8 epidemiology working group (BTV-8WG) was established by EFSA on 6th of October 2006.

This report describes the results of the different tasks that were performed to investigate factors associated with introduction, establishment and spread of BTV-8 in north-western Europe. They concern the following aspects of the outbreak investigation:

• The introduction of the BTV-8 serotype (focussing on the place, time and possible routes for this introduction); 
•  A section on clinical aspects in which the nature and severity of the disease caused by this strain are described; 
• The characterisation of within-herd spread to assess among others the ratio between sub-clinical and clinical cases. This has implications for detection of BTV-8-infected herds; 
•  The information on factors favouring virus establishment. This includes the results of vector surveillance in the affected countries and factors that can affect virus persistence;
• Elements that may influence short-distance spread. This included the study of observed speed of local spread and the characteristics and pertinence of the 20 km zones;
• Finally, in the section on factors affecting long-distance spread into new areas, the potential for vector spread through wind and restrictions on animal movements were considered.
  

The provisional results, to be delivered by 31 January 2007, were presented to the BTV-8WG on the 6th of February 2007 and to the European Commission and Chief Veterinary Officers (CVOs) at an EC meeting on the 7th of February 2007 in Brussels. Taking into account the feedback from the meetings on the 6th and the 7th of February 2007 the reports were then revised. For the final report additional data through 31 January 2007 were considered as well, where relevant.

A full epidemiological analysis of the relationships between disease incidence and possible risk factors requires a multivariable analysis that allows taking account of possible interactions and confounding between important risk factors. Due to time constraints it was not possible to conduct such an analysis. Thus, the main focus of this report is on the descriptive and exploratory analysis of the epidemic.

The main findings reported by the EFSA bluetongue working group are reported below.

1. Statistical modelling showed that the initial infection occurred in the area close to Maastricht. Difficulties in the initial diagnosis were due to the fact that the disease had never been detected in the area before.

The source of the introduction of BTV-8 could not be identified and the exact origin and route of the introduction of BTV-8 thus far remains unknown. However, the absence of legal import of ruminants from outside the EU into the Area of First Infection and the absence of BTV-8 from southern Europe suggest that the introduction of the BTV-8 infection into north-western Europe is likely to have occurred via a other than through import of infected ruminants. Hence, the potential for introduction via mechanisms other than those incriminated in previous introductions also needs to be considered. Specifically, the potential for Culicoides to be imported along with or independently of the import of animals, plants or other ‘materials’ merits further study.

2. In BTV-8 infected herds, cattle and sheep expressed different clinical signs. BTV-8 associated clinical signs were much more common in sheep than in cattle, although it was found that a small number of cattle within a herd can show distinct clinical signs.

3. Based on the (sparse) data from whole herd sampling there was a trend suggesting a high proportion of cattle to be PCR- and sero-positive in infected cattle herds and a small proportion of sheep to be PCR- and sero-positive in infected sheep herds.

Taking into account the findings in relation to the within-herd patterns of clinical signs and serology, it can therefore be suggested that:

• for sheep flocks a monitoring system based on clinical signs should be considered; whereas 
• for cattle a monitoring system based on serological surveillance appears to be the more effective approach.
  

4. The BTV-8 virus was found to be present in vectors (Culicoides species) which are endemic to north-western Europe. C. imicola, which is thought to be responsible for at least 90% of BTV transmission in the Mediterranean Basin, was not found amongst a total of approximately 100,000 Culicoides collected in the infected MS. This demonstrates that species endemic to the palaearctic region are capable of transmitting BTV and – judging from the rapid spread of the virus – no pre-adaptive phase was required in the indigenous Culicoides. Species found to be PCR-positive were C. dewulfi (a species breeding exclusively in the dung of cattle and horses) and C. obsoletus / C. scoticus.

Vectors of BTV in other parts of the world have been shown to transmit a range of other viral pathogens of livestock (African horse sickness virus, Akabane virus, epizootic haemorrhagic virus, equine encephalosis) suggesting that such pathogens might be transmitted if they were to be introduced into northern Europe during climatically favourable periods. However, overall, there is a paucity of information on the behavioural activities of vector species of Culicoides, especially in relation to their interactions with host animals and their biting activities. Therefore, detailed data are urgently required before clear and reliable recommendations can be provided to the veterinary authorities on the subject.

Studies in France and in the Netherlands showed that later in the season, when the temperatures began to drop, more Culicoides were captured inside rather than outside stables. This undermines the rationale of the recommendation to house livestock indoors to reduce the Culicoides attack rate and indicates a need to revisit the definition of vector-free period to be when “…<10 Culicoides are found in a light trap suspended outdoors for one night …”. More comprehensive data on the effectiveness of control measures to prevent influx of Culicoides into stables are needed to allow meaningful recommendations.

Low numbers of adult Culicoides principally of the Obsoletus Complex, including freshly blood fed individuals, were on occasion captured in light traps operated throughout the winter (January, February and March 2007) in various MSs in north-western Europe. In all likelihood this persistent activity of adult Culicoides owes much to the mild temperatures that have continued to prevail across northern Europe during the winter of 2006/2007. However, the numbers of midges remaining active appear to have been too few to sustain the BTV transmission cycle as no sero-conversions in sentinel cattle was reported during that period.

Within the infected area, the presence of BTV-8 was favoured in locations that were warmer on average, where temperatures varied less throughout the year and where temperatures rose quickly in spring reaching a peak earlier in the year. The daily variation of number of Culicoides captured in the Netherlands was found to be linked with prevailing temperatures, for all Culicoides species present. The prevailing high temperatures in the summer of 2006 resulted in high numbers of various Culicoides species. Comparing current meteorological data in Germany to historical long term weather records (from 1961 until 1990) it was found that the adjusted mean temperature were exceptionally high during the months of July and September – October 2006.
Thus, warm climatic conditions may favour the establishment of these viruses after they have been adventitiously introduced in a new part of Europe. The number of new bluetongue cases over time was equally influenced by changes in temperatures. The lag time between a change in temperature and a correlated change in number of outbreaks was estimated to be about 4 weeks. This resulted in an initial peak and then a second peak of new cases which were separated by a cooler period.
Besides temperature, the number of observed bluetongue cases was also related to other environmental factors such as altitude and animal density.

5. Local spread was modelled and found to occur at a rate of about 2 km per day or approximately 15 km per week. This estimation is consistent with published data on active flight distances covered by Culicoides and observed spread of BTV-outbreaks in Sardinia.

6. It was demonstrated that wind may affect spread over long distances. In particular, the density of the observed wind events contributed to explaining, at least in part,

• the predominant east-west spread of the epidemic, 
• the more limited spread towards the north and south, and 
• the absence of recorded outbreaks in the U.K.

The movement of cattle mainly occurred in a north-western direction and the extension of the epidemic to the east can therefore not be explained by these transports.

In conclusion, changes in climatic conditions coupled with increased worldwide traffic might increase the risk in the appearance and the establishment of diseases in parts of Europe that were thus far exotic to those regions.