The Scottish Government

« Previous | Contents |

Listen

Introduction

Bluetongue virus ( BTV) causes significant disease losses among ruminant and camelid livestock. In recent years, the geographic area affected by BTV has increased significantly with disease now occurring in animals across an ever increasing area of mainland Europe with the virus somehow surviving over winter in Northern Europe in 2006/2007. BTV is carried by midge vectors and a strain of this virus, BTV strain 8 ( BTV 8), was detected for the first time in England in the autumn of 2007.

There is a high likelihood that BTV 8 will enter Scotland in the foreseeable future but there is considerable uncertainty about many aspects of how this disease will behave in the GB environment. At this stage there is a limited understanding of how both Scottish livestock and Scottish midge populations will respond to BTV 8 as well as how effective the existing disease control measures might be. Potential control measures include midge (vector) control, BTV 8 vaccination and ruminant movement restrictions combined with surveillance for early detection . Slaughter, while also an option, is unlikely to be widely used due to the midge-borne nature of the disease.

Despite the knowledge gaps there is a current and urgent need to determine the best control strategies for Scotland. Prior to implementation, the effectiveness of a range of potential control options must be evaluated in order to prepare for any incursion of BTV. Epidemiological modelling was considered to be the most suitable approach to help integrate the emerging biological information and expert derived knowledge about how BTV will most likely behave in Scotland. In turn this modelled output provided the basis for this report's economic analysis. The four objectives specified in the project specification were:

Objective 1: Development of feasible incursion scenarios

Objective 2: Development of epidemiological scenarios

Objective 3: To develop an economic consequences model for the consequences of BTV 8 spreading to Scotland

Objective 4: To carry out economic evaluation for the agreed control strategies for each Incursion Scenario

Given time constraints the study was restricted to a limited number of possible incursion scenarios and a limited range of control options, representative of a range of scenarios that could occur. The range of modelling tools adopted for this research allowed us to pull together existing data and utilise it in the best way to meet the project objectives. The expert panel helped with estimates where data were missing and generally advised upon procedures.

Research Methods

A multidisciplinary expert panel was formed, including BTV and midge experts and disease-control policy makers. The panel agreed a range of feasible BTV incursion scenarios, patterns of disease spread and specific control strategies to investigate. Our study utilised data already held by different members of the project team but was predominantly desk based, applying quantitative methodologies with pre-existing epidemiological models. We explored the most likely distribution of the disease given Scotland's agricultural systems, unique landscape and climate. We engaged with Scottish Government officials and with livestock industry representatives to help inform decision making and prioritisation of disease control options should BTV spread to Scotland.

Key Findings

1. Development of feasible Incursion Scenarios (Objective 1)

The Incursion Scenarios agreed by the expert panel and the Scottish Government were:

1. northwards spread, with BTV arriving in April 2009
2. northwards spread, with BTV arriving in July 2008
3. northwards spread, with BTV arriving in September 2008
4. import of infected animals in April 2009
5. import of infected animals in September 2008

Subsequently, a limited range of Control Strategies was agreed with the expert panel. The impacts of the following Control Strategies on the incursion scenarios above were investigated:

C1) implementing the minimal requirements (movement restrictions) only

C2) vaccinating 100% of holdings in a border protection zone ( PZ)

C3) vaccinating 80% of holdings in a PZ to the Highland Boundary Fault (B/F) line

C4) vaccinating 50% of holdings in a PZ comprising the whole of Scotland

C5) vaccinating 80% of holdings in a 100km PZ around the first identified holding (only applied if the incursion occurs above the Highland B/F line).

C6) vaccinating 80% of holdings in a PZ comprising the whole of Scotland (an additional piece of work commissioned in June 2008 at the request of SG).

2. Development of epidemiological scenarios (Objective 2)

The outputs from our epidemiological models indicated that for most incursion scenarios infection seldom spreads after the initial incursion. For incursions due to Northwards spread, only an incursion in July (Incursion Scenario a) resulted in outbreaks becoming more widespread in a substantial number of modelling simulations. However, if undetected, an import of infected animals is likely to result in a large-scale outbreak, regardless of the time of the year when the import occurs.

For all incursion scenarios, vaccination is efficient at controlling the spread of BTV and widespread outbreaks will usually be prevented in areas using geographical targeted vaccination with high levels of vaccine uptake by farms. In the face of random importation of infected animals, a higher level of spatial coverage at a lower level of vaccine uptake by farmers (assuming this was evenly distributed through the livestock population) is usually more effective for disease control than a higher level of uptake in a small (and possibly wrong) area. However, the optimum strategy for reducing disease spread is a high level of uptake over a wide area. Vaccination also has a marked impact on the longer-term dynamics of BTV. Where vaccination is used infection typically dies out within two years (See Figure 1), whereas it persists where only minimal control measures are applied. It is important to note that 100% efficacy of the vaccine was assumed in the absence of alternative data.

As part of this project the project team included recent research on competent vector (midge) distributions in Scotland. The analysis of Scottish vector data generated maps of suitable habitat for the bog-heathland Scottish biting midge Culicoides impunctatus. The Northern uplands are at high risk of supporting large C. impunctatus populations. The Scottish biting midge is most likely to overlap with farmland, and with domestic ruminants and farm associated vectors in the North West Highlands, along the Great Glen, at the foot of the Grampians and in the Scottish Borders. Domestic ruminants overlap with large populations of red deer on the Cairngorm plateau, along the Moray coast and sporadically through Highland areas. It is not yet possible to predict how the numbers of farm-associated midge vectors ( C. obsoletus or C. pulicaris complexes) vary across Scotland on the basis of current vector surveillance data, but an outline for a future framework for such predictions formed part of this report.

3. Economics (Objectives 3 & 4)

The objectives of the economic aspects of the work were two fold:

• To develop an economic consequences model for identifying, measuring and valuing direct and indirect socio-economic consequences (costs due to disease control and other consequences) of the virus spreading to Scotland.
• To conduct, under each of the incursion scenarios, an economic evaluation of the strategies available for controlling the disease.

The results for both these economic objectives have been combined for simplicity:

Our economic consequences model was based upon evaluating the size of direct and indirect costs arising from an incursion of BTV into Scotland alongside the cost of undertaking prevention measures. The most efficient control strategy is therefore the one that results in the lowest BTV outbreak losses occurring.

With our existing knowledge base it is not yet possible to estimate the probability of each incursion scenario evaluated. The selected scenarios are but a few of many possible incursions that are not mutually exclusive. This means that our BTV 8 control options must be compared within each specific incursion scenario.

For most Incursion Scenarios the Control Strategies of vaccinating 80% (C6) and 50% (C4) of holdings in a protection zone comprising the whole of Scotland were best (i.e. gave the lowest BTV outbreak losses of all the vaccination strategies). The only exception was for importation of infected animals in September 2008 (Incursion Scenario e) where vaccinating according to the location of the outbreak (Control Strategy C5) was best. The highest BTV outbreak losses are always associated with vaccinating 100% of holdings in a Scottish Border PZ (Control Strategy 2). Under some circumstances, the no vaccination option (C1) delivered the lowest outbreak losses. However, given the uncertainties surrounding the probabilities for each Incursion Scenario and the relatively small differences compared with the vaccination options; control by vaccination is the recommended approach.

It was concluded that no particular economic advantage or disadvantage was associated with raising uptake of vaccine from 50% (C4) to 80% (C6) in a PZ for the whole of Scotland. It is assumed that the value of the economic benefits of the higher rate of vaccination uptake arising from reduced spatial spread of disease (highlighted in the epidemiological results) must have been offset by the greater costs in achieving such increased uptake. In this case and given the uncertainties inherent in this type of predictive analysis, targeting the higher rate of vaccine uptake (C6) might be considered prudent.

The separate potential impacts of BTV 8 incursion on the sheep and cattle sectors were compared using two examples: Incursion Scenario (a) with Control Strategy 4 and Incursion Scenario (a) with Control Strategy 6. Direct losses due to an outbreak of BTV (mortality, morbidity, veterinary costs etc.) were greater for sheep than cattle, since infected sheep tend to express more symptoms than infected cattle. However, these outbreak losses were dwarfed by other costs (direct and indirect), which were dominated by cattle associated losses. This result must be emphasised through communication with the Scottish cattle sector. Our estimates for total direct costs for a Scottish BTV 8 outbreak were comparable with recently published estimates from the BTV epidemics in the Netherlands (about £30m per annum). However, direct costs are likely to be much smaller than indirect costs (loss of markets, price effects etc.).

Our study results estimate that the indirect costs resulting from a BTV 8 outbreak may exceed £70m per annum. This reinforces the importance of investment in baseline costs that reduce the risk and extent of any potential incursion. (We must, however, emphasise that indirect costs are extremely difficult to estimate with the current knowledge base). Indirect cost estimates were dominated by reduced demand for beef and hence lower beef prices. The extent of this effect will depend on consumer reaction to news of a BTV outbreak in Scotland. This is very difficult to predict but was assumed to be small at -£0.25/kg (since there are no obvious public health implications). However, a small reaction is magnified by the sensitivity of the beef price to a small demand change (elasticity) and the large quantity of beef produced in Scotland.

The current costs of surveillance and other related activities that are aimed at reducing the risk of incursion and/or limiting the damage of any incursion have also been considered. The current baseline costs are estimated to be £141m in present value terms over the 5-year time horizon considered within this study. Importantly the benefits of avoiding disease incursion exceeded current baseline costs of prevention in all scenarios evaluated. This indicates that the current baseline costs are fully justified. However, without more information about the effectiveness of baseline costs in each scenario it was not possible to investigate this aspect in more detail.

To ensure that our estimates remain robust we carried out a sensitivity analysis of the economic models. Control Strategy (Vaccination) results are little affected by:

• Variations of up to 5% in the main assumptions used in the economic model,
• Taking worst (95th percentile) or best (5th percentile) epidemiological predictions,
• Whether or not a licence was available for movement to slaughter and
• Whether the outbreak continued unabated or tended to decline from year 3 to year 5.

Such robustness is reassuring. However, great uncertainty still surrounds the probability and nature of incursions of BTV into Scotland and the relative economic efficiency of alternative prevention and control options.

In our study we also explored the extreme ends of possible epidemiological model outputs. Very minor or very major BTV 8 outbreaks and associated disease loss estimates made little difference to the overall economic assessments for the full range of alternative Incursion Scenario/Control Strategy combinations, already presented, based on the average epidemiological outcomes. This combined with the results of the sensitivity analysis and the consistency between incursion scenarios is reassuring as it suggests that the choice of the best control option is robust to the nature and extent of the incursion.

Issues for Future Consideration

Our knowledge and information base for bluetongue infection as occurring in Northern Europe is increasing all the time and new information is placed in ANNEX 5 of the full technical report available on the Scottish Government website. Vaccination was assumed to be 100% effective for the purposes of this study but the true efficacy has yet to be determined. In addition, the project focused on BTV serotype 8 however the UK should now be considered at risk from other strains of BTV such as BTV 1. Vaccination against BTV 8 will not protect against other strains such as BTV 1. This is discussed in ANNEX 5 (a) of the full technical report. Revised information about changes in the currently available tests and the GB capability to provide accurate test results in the face of widespread BTV 8 disease is provided in ANNEX 5 (b) of the full technical report. During the last weeks of the project evidence for the risk of horizontal and vertical transmission of BTV came to light. The economic modelling does not incorporate all this information at this time but it is vital that this emerging information is taken into account when making decisions about how best to control bluetongue in Scotland; this is discussed in ANNEX 5 (c) of the full technical report.

Conclusions

If a BTV outbreak was to occur in Scotland, the costs are estimated to be £100m per annum (£30m in direct costs and £70m in indirect costs). Although the symptoms of disease are expressed more often in sheep than in cattle, the losses to the cattle sector from a BTV 8 incursion into Scotland would exceed those to the sheep sector.

This research project has concluded that the cost of surveillance and disease control through vaccination are both justified in the face of a probable BTV outbreak.

Vaccinating 80% (C6) and 50% (C4) of holdings in a PZ comprising the whole of Scotland were the best Control Strategies (i.e. gave the lowest mean BTV 8 outbreak losses of all the vaccination strategies). Since the higher rate of vaccination uptake offered the greatest reduction in the spatial spread of BTV and given the uncertainties inherent in this type of predictive analysis, targeting the higher rate of vaccine uptake (C6) might be considered prudent.

Notes

This research was funded by the Scottish Government Rural and Environment Research and Analysis Directorate ( RERAD), Contract Research Fund. Commission Number: CR/2007/56, 2008. The full technical report is available on the Scottish Government website.

This research was conducted by SAC on behalf of EPIC

Figure 1: Computer simulation of BTV8 spread under Control Scenario 6 (vaccinate 80% farms in a PZ comprising the whole of Scotland). Figures a to e represent the spatial extent of the infection under the five possible incursion scenarios (a-midge April.09, b-midge July 08, c-midge Sept. 08, d-import April 09, e-import Sept. 09). The legend indicates the proportion (%) of modelling simulations (replicates) for which each 5km grid square had at least one affected holding.

« Previous | Contents |