Thursday, October 11, 2012

Defra's Calculator

Last month, Defra produced an astonishing news release which trumpeted the following information:
There has been a notable decrease in the incidence rate over the past 6 months, mainly as a result of an increased number of tests on unrestricted herds compared to last year. The AHVLA are currently carrying out a data checking exercise to validate these test statistics.
The provisional June 2012 incidence rate is 4.2%, compared to 6.0% in June 2011. However, care needs to be taken not to read too much into short term figures, especially as this figure includes a number of unclassified incidents. As such, the incidence rates are subject to further revisions as more tests and their results for the period are input.
The number of new herd incidents during the period January to June 2012 was 2,706 compared to 2,720 for January to June 2011. The number of tests on officially TB free herds was 41,656 in January to June 2012, compared to 34,667 in January to June 2011.
The number of cattle compulsorily slaughtered as reactors or direct contacts was 18,213 in January to June 2012, compared to 18,081 in January to June 2011.
So all good news then?
Well, that depends on who is punching the buttons. We have archived the key figures for Jan - June 2011 and guess what? They do not compute to this conclusion at all. In the same 6 months of 2011, 6073 herds were under restriction 'due to a TB incident'. And that was 7.35 per cent of the total herds registered in GB. This year, 2012  Defra's current spread sheet ( apologies -  can't get a link to stick) shows 6567 herds in that position, but with fewer herds on the database, that equates to 8.2 per cent.

Cattle slaughtered are up too. The totals being 18,082 in 2011 compared with 18,645 so far this year. An increase of 563 animals. (Defra's figure in their announcement ignored 432 IRs, but they are still dead and for accurate comparison, IRs were counted in the total last year)

New herd incidents? That is benchmark Defra usually present, as it lower by a a mile than the tail end of herds under restriction and 60 day testing for years. That figure is pretty much the same at just over 2700.

So where has this alleged drop of almost 2 percent come from?  June 2011 was the last time Defra produced its monthly stats on the old system, and then dear old SAM took over and the number of tests logged increased by 30 percent. In fact his tally is 49,528 tests for GB in just six months. In some areas more than one test per registered herd in just six months.

 Defra explained this in a previous news release, saying that they were now counting pre and post movement tests, and also removing gamma failures and slaughterhouse cases from some totals as well, so figures may not be comparable. But despite all this tinkering, the number of herds under restriction for a 'TB incident'  is up, cattle slaughterings are up, and new herd breakdowns are at best, static. And yet Defra say that as a percentage,  new herd incidence is down? Something here doesn't ring true, but hey, the computer has spoken.

Could it be that those running the SAM system are refugees from this train wreck of computer generated mathematics?

UPDATE : We see our furry little friends in the Badger Trust have issued a press statement for 'Immediate release' trumpeting that Defra report a '30 per cent drop in cattle TB' - so stop the cull.
Oh dear. Another calculator which can't add up, and an organisation which can't see beyond a skewed headline.

Further investigation into Defra's stats do not support the 'wider testing' argument - should they make it.
The West region has recorded at least one test per registered herd - in just 6 months. And it's almost the same in Wales. So, in annual testing areas, with preMT ,  yes a huge increase in tests - but as Defra say, dear old SAM is now including ALL these tests which wasn't the case before.  And they haven't discovered many more reactors than last year. But the tail on this particular dog gets longer, as herds under restriction are not cleared by removing sentinel tested cattle. So to reiterate, and just for Uncle Jack Reedy, the key figures for the 6 month period Jan- July 2012 are here:
Herds under restriction during the period  - 6567, up by 494, almost 1 per cent.

Cattle slaughtered during the period - 18,645 up by 563.

New herd breakdowns - static at just over 2700.

Tests carried out, particularly in the West and Wales - up 30 per cent.

So that's a  30 per cent drop in cattle TB? If you say so, Jack. If you say so.

We don't think so.

No comments: