Modelling - SOAs
In the 2001 FMD epidemic, two factors resulted in the exceptional difficulties associated with control. First, the rapid rise in numbers of holdings affected meant that logistical capability was strained to capacity and beyond, and second, the wide geographical extent meant that the control effort could not be concentrated in one area, further increasing the cost of control [9]. Here, we investigate the impact of exemptions on both the number of premises affected, and the geographical distribution.
Because of the potentially complicated interactions between spatial spread, spread due to different species and spread due to the dynamic network of livestock movements, we use a simulation approach to summarise the impact of all these factors on the two outcomes described above – number of premises and geographical spread. In our previous work [3], CTS and AMLS movement data were used to construct an individual holding based model of the initial spread of FMD in GB to determine the susceptibility of the GB livestock industry to future outbreaks. Transmission through movements was modelled, with additional local spread unrelated to the known movements, as well as intra-SOA spread. Simulations showed that movements can result in a large nationwide epidemic, but only if cattle are heavily involved, or the epidemic occurs in late summer or early autumn. Inclusion of local spread can considerably increase epidemic size, but has only a small impact on the spatial extent of the disease. Importantly, the inclusion of intra-SOA spread had only a minimal impact on final epidemic size. However, the increased number of SOAs since then and the additional effect of CTS Links merits further investigation of their impact on FMD spread.
Therefore, this approach [3] was updated to consider the updated data on both SOAs and CTS Links and was re-run with the latest movement data from AMLS and CTS for 2008; the model was run for England and Wales only, as updated Scottish data were unavailable at the time of the analysis. We also consider two stocking scenarios for intra-SOA spread: (1) where the recorded distribution of stock numbers within SOA member holdings is correct; and (2) where stock are evenly distributed amongst SOA member holdings. While the mechanisms of SOAs and CTS Links are different, for this study we assume that their role in FMD spread is similar – rapid spread amongst holdings that would otherwise be of limited risk to each other, thereby increasing the potential for onward spread to other holdings. Although this is a worst-case scenario, there is currently insufficient information on SOA stock levels as well as intra-SOA and intra-Chain movements to develop realistic alternatives.
We allow the epidemic to run for a period of one month, in order to explore the complex relationship between the explicit network of livestock movements, and the possible role of dispensations on disease spread. This is not meant to be realistic; identification of FMD on a holding would result in an immediate movement ban, possibly at the national level, and the probability of detection is related to a number of factors, most importantly the number of premises affected. However, use of a fixed time frame for spread provides for a basis of comparison with our previous work [5] and is an indication of the relative impact of these dispensations on FMD spread, compared to 2001 when silent spread occurred for roughly this period.
As can be seen in Figure 7, when intra-SOA spread is limited to only holdings with reported stock, the inclusion of SOAs appears to have little effect on epidemic size. This is consistent with our previous results [3] which did not investigate alternative stocking scenarios. As the majority of SOAs have reported stock at only one holding within the SOA, there is little intra-SOA spread recorded in the model as there are no other stock containing holdings within the SOA to infect with the disease and subsequently spread out from via local spread.