I've blogged about this when I saw Christl Donnelly's talk at the UCL Symposium, last year (in a nutshell, the idea is that badgers are potentially responsible for infecting cattle with TB; thus "reducing the chance of contact" between cattle and badgers would reduce the number of infected cows. NB: Cows infected with TB need to be slaughtered, which also has a clear economic impact).
Apparently, the UK government has now decided to pilot badger cull in selected areas of England. But: despite the best effort of the statisticians involved (and Christl was very clear in explaining how difficult was for them to design the trial), the evidence was still not completely conclusive as to whether the cull may produce benefits.
Last year, the government delayed the decision in order to sort out licensing conditions, which are now met. But, as far as I can see, in the meantime nothing has changed in terms of the available evidence. The RSPCA are vehemently protesting, arguing that the way forward is vaccination for badgers (and, in the future, when the bovine vaccine is ready, for the cattle as well), not culling.
Of course I'm no expert, and the issue is quite controversial. But perhaps, given the inconclusive level of the statistical evidence available, it would be good to trial both possible strategies and re-assess the decision-making process when more data on the impact of vaccination are available.
I think I wrote it in my original post already, but I can't help but thinking that this problem would benefit from a proper decision analysis complemented by the evaluation of the expected value of information.
Please comment on the article here: Gianluca Baio's blog