What is the most cost effective path to TB freedom?
The current national bovine tuberculosis (TB) control strategy focuses on reducing and maintaining low possum densities in areas with infected livestock, and the testing and movement control of cattle. This strategy has successfully reduced the number of infected cattle herds from over 1,400 in 1994 to fewer than 90 now. The Animal Health Board now proposes to eradicate TB from all wild animal populations over 2.5 million hectares of New Zealand by 2026. This ambitious but achievable goal will depend on a ‘proof of freedom’ framework designed by Landcare Research. This framework provides an objective assessment of the probability that TB has been eliminated from an area (PNoTB), based on recent control of wildlife vectors of the disease and lack of detection of TB in livestock, possums, or other wildlife sentinel species.
Under this framework, an area will be declared TB-vector free when PNoTB is greater than some predetermined risk-management threshold (e.g. 95% probability that TB has been eliminated) that represents a low risk of operational failure (i.e. a 5% chance that TB is still present in wildlife vectors). When PNoTB is still below the threshold, undertaking further possum control will increase the probability of achieving eradication. However, TB may have already been eradicated, and all that is needed to prove this is additional wild animal surveys. Obviously a key need is to know which option to choose for any specified area. For example, should funding be used to kill more possums or to survey sentinel species such as wild pigs and deer for TB? Both strategies will help increase PNoTB, but which would be most cost efficient?
To explore these complex resource-allocation questions, Graham Nugent, Pen Holland and colleagues are using the TB-freedom problem outlined above as the initial, and arguably one of the simplest, of several case studies (including island and disease eradication problems) to construct whole-system models to simulate all the key components of the Resource Allocation Framework (RAF) outlined in Fig.
Data from the Blythe Valley (North Canterbury) is used here to illustrate how such a RAF will help TB managers. TB emerged in cattle herds in the Blythe Valley in the 1990s, and possum control began in 2000. By 2004, annual ground control of possums had reduced the number of infected herds to low levels but TB was still present in cattle. Intensive annual possum trapping continued until 2008 when just 16 possums were caught from 6105 trap nights. Simulating this possum control history in a model of TB epidemiology in possums predicted a 99% likelihood that TB had been eliminated from possums as early as 2005.
Despite that prediction, TB was still present in cattle in 2005, making it prudent to assume there was still a 50:50 chance that infection was present in wildlife populations of either possums or ferrets. However, using data from trapping and from necropsies of possums and ferrets undertaken over the next three years, the Proof of Freedom Framework predicted a >95% likelihood that TB had been eradicated by 2008. One of the most important factors contributing to this high probability of eradication was the large amount of data from traps that caught no possums. If there is a zero catch rate at a trap site, there is a near-zero chance TB could persist within a 200–300 m radius of that site, because there are too few possums to maintain the disease.
Since 2007, no infected cattle have been found in the Blythe Valley. With hindsight, the last few cases of TB in cattle in 2005–2007 were probably animals that didn’t react to the TB test or were due to within-herd infection. If so, TB was probably eradicated from possums by 2005 (as predicted by the Possum-TB model).
However, a further $256,000 was spent on pest control, arguably for no benefit if TB had already been eliminated from wild animals. The obvious difficulty is that managers had no way of confirming that TB had been eliminated, so their decision to continue precautionary control was sensible. Nevertheless, some of the expenditure could have been saved if there had been an earlier shift away from simply killing possums toward both killing possums and using the wildlife surveillance data to ‘prove’ TB was absent.
The aim is therefore to develop whole-system models that will enable TB managers to allocate funding among control and surveillance strategies, in order to maximise the rate of TB eradication and to provide timely confirmation of success. Graham and his colleagues consider that a whole-system model for TB will improve the allocation of control and surveillance resources and thereby improve eradication efficiency.
This work is funded by the Foundation for Research, Science and Technology (Programme C09X1008: TB and Multi-Pest Suppression Systems).
