Resumen de Editorial: Bovine tuberculosis:: how likely is a skin test reactor to be uninfected?
ACCURATE diagnosis of bovine tuberculosis (TB: infection with Mycobacterium bovis) in live animals is difficult. A large part of the problem is due to the diagnostic tests being imperfect, and the comparative skin test for diagnosing TB in cattle is no exception. Focus is often put on the likelihood that a diagnostic test will detect an infected animal as positive – the test sensitivity – which for the comparative skin test may be as low as about 50 per cent (Downs and others 2011). While this means that many individual infected cattle may go undetected, infection is far more likely to be detected at the herd level.⇓ There is another side of diagnostic test performance which needs considering and that is test specificity. Specificity is the probability that an uninfected animal will test negative. One way to estimate this would be to test a large number of uninfected cows and see how many negative results you get, but this is not straightforward. For instance, how do you ensure that the animals you test are truly uninfected given that no perfect test exists? How do you find sufficient numbers of uninfected animals to test? And, perhaps most important of all, how do …
