How much do badgers contribute to bovine TB in cattle?
There is overwhelming evidence that badgers are involved in the transmission of bovine TB to cattle. What is much less clear is ‘how much’ of the problem is down to badgers, relative to other factors like the movement of infected cattle or the inaccuracy of the TB tests. One study which has attempted to quantify the contribution of badgers to bovine TB in cattle is this study by Donnelly and Nouvellet (2013).
It is a study which is often mentioned in discussions about bovine TB and badgers. As with other studies and statistics you often hear if from both ‘sides’ either as a justification that badgers are a significant part of the problem…. “badgers cause half of breakdowns”….or badgers are only a tiny part of the problem….. “only 6% of breakdowns are due to badgers”. To try and add some clarity I have produced a one page summary below. If you are interested I have added more detail on the methods below the summary, or alternatively click on the links to the original paper above.
A bit more detail on the study methodology
The paper uses a mathematical model based on an early paper by Donnelly and Hone (2010). The model involves a series of equations. Several parameters in these equations are already known or can be estimated from other data sources (on the left of the figure below). Other unknown parameters are estimated using the model itself, essentially by solving the equations to find the missing values. See a list of model parameters below.
The authors use the data produced by this model to then produce estimates for two key figures
1) The overall % contribution of badgers to cattle TB incidence
2) The % of transmission which is badgers to cattle.
You may be thinking…”aren’t they the same thing!!??” Not quite. We know that there is some degree of cattle to cattle transmission. This means that herds directly infected by badgers can potentially spread the disease further by the cattle to cattle route. Or to turn that round, if we can estimate the change in TB in cattle following badger culling (say 50%) then the % which is direct badger to cattle must be less than that (ie less than 50%) as whatever rate of TB comes from badgers will then be amplified by the spread through the cattle population. For example, if you imagine that badgers infect 5 herds directly, but cattle from these herds then transmit the infection to a further 10 herds (these are just rough numbers for an example), then badgers would contribute to TB in 15 herds in total (5+10), but only 5 directly. The model in the paper uses the data available to estimate the rate of cattle to cattle and badger to cattle transmission that are most likely, or will best fit the data.
As with any model there are limitations and assumptions and the values produced are estimates with a degree of uncertainty. This model uses data from the RBCT areas, so it is representative of those areas at that time. The contribution is likely to be different in parts of the country such as the low risk area and potentially different in other areas too, such as parts of the HRA where the cattle or badger population differs from the RBCT areas.
The model also assumes that the cattle to badger transmission is ‘negligible’, clearly some cattle to badger transmission does occur, but again, how much is unclear. As TB in badgers increased following foot and mouth (during which there was no TB testing) the authors of this study also ran the model excluding the initial proactive culls undertaken in 2002 after the 2001 epidemic (so only using 7/10 triplets). However, the exclusion of this data increased the uncertainty in the model.
The paper also reports the results from an earlier analysis by Jenkins et al. (2010) which looked at changes in TB incidence following the proactive culls in the RBCT. This can be viewed as another estimate of the overall contribution of badgers to cattle TB, as this is the level of reduction achieved if the badger-to-cattle route is reduced (by culling the badger population).
What do the results mean?
Firstly it is worth stressing that there is a large amount of uncertainty around the estimates produced, particularly the overall contribution of badgers as estimated by the model. The confidence intervals for this estimate are from 9% to 100%, with the 52% figure essentially in the middle of this. This is likely to reflect the uncertainty around the numbers/model itself, but also the fact that the data comes from 10 different RBCT areas, which themselves will vary. As stated above the data are specifically from the RBCT areas and different rates of transmission may occur in other areas. As such, the exact values quoted should be treated with a note of caution.
Both the analyses support the conclusion that badgers play a major role in maintaining M. bovis in cattle as they contribute to a significant % of breakdowns and culling (ie removing or reducing the badger source) can lead to significant declines within cull areas. The results of the model also support the conclusion that cattle to cattle transmission is important in disease transmission as it is responsible for roughly half of breakdowns directly (ie the other 50% not caused by badgers) AND because it amplifies the impact of transmission from badgers, further spreading TB through the cattle population. If accurate, these results suggest that reducing badger to cattle transmission at a proportion of farms could potentially lead to significant benefits in TB reduction. It also suggests that any cattle measures which minimise cattle to cattle spread could also have significant benefits.