Reader Comments

Post a new comment on this article

Analysis of further data (to 25 February 2011) on the impacts on cattle TB incidence of repeated badger culling

Posted by ChristlDonnelly on 11 Apr 2011 at 13:50 GMT

Analysis of further data (to 25 February 2011) on the impacts on cattle TB incidence of repeated badger culling

Christl A. Donnelly1*, Helen E. Jenkins1, Rosie Woodroffe2
1 Department of Infectious Disease Epidemiology, Imperial College London, London, United Kingdom
2 Institute of Zoology, London, United Kingdom

* E-mail: c.donnelly@imperial.ac.uk

Since publication of the paper “The duration of the effects of repeated widespread badger culling on cattle TB following the cessation of culling” an additional 19.5 months of cattle testing data have become available. These allowed analyses to be updated, following updates published as comments in May and July 2010.

In the time period from one year after the last proactive cull to 25 February 2011 (the post-trial period), the incidence of confirmed breakdowns in the proactive culling trial areas was 31.5% lower (95% CI: 19.1% to 42.0% lower) than in survey-only areas, and on lands up to 2km outside proactive trial areas was 4.4% lower (95% CI: 27.4% lower to 26.0% higher) than outside survey-only areas.

Exploratory analyses stratified by 6-month periods (Table 1) are consistent with an ongoing benefit of proactive culling continuing through the latest 6-month period analysed (49 to 54 months post-trial).

The effects observed outside trial areas remained consistent with no ongoing effects of proactive culling in these areas.

The post-trial results must, of course, be considered in the context of the smaller reduction seen inside proactive trial areas and the increased incidence seen outside proactive trial areas in the period from the end of the initial proactive cull until one year after the last proactive cull in each triplet. From the start of the RBCT to 25 February 2011, incidence of confirmed breakdowns in proactive culling areas was 27.4% lower (95% CI: 20.2% to 33.9% lower) than in survey-only areas. In areas up to 2km outside the trial area boundary of proactive culling areas, incidence of confirmed breakdowns was 8.3% higher (95% CI: 14.6% lower to 37.4% higher) than in areas up to 2km outside survey-only areas.

Table 1 Estimated effects of proactive culling on the incidence of confirmed cattle TB breakdowns inside trial areas. Analyses adjust for triplet, baseline herds, and historic TB incidence (over three years). Results are split by 6-month period post-trial and include breakdowns from one year after the last proactive cull to 25 February 2011.

Time period (post-trial) Estimate (95% confidence interval) p-value
-------------------------------------------------------------------------------------------------------
Months 1-6 -51.5% (-70.7%, -19.6%) 0.005
Months 7-12 -39.6% (-62.7%, -2.4%) 0.040
Months 13-18 -48.6% (-67.1%, -19.9%) 0.003
Months 19-24 -25.8% (-50.7%, 11.6%) 0.15
Months 25-30 -29.5% (-55.2%, 11.1%) 0.13
Months 31-36 -9.1% (-41.9%, 42.2%) 0.67
Months 37-42 -29.9% (-54.1%, 7.1%) 0.10
Months 43-48 -27.6% (-54.9%, 16.2%) 0.18
Months 49-54* -37.0% (-62.4%, 5.4%) 0.079

[* This time period had 4.6 triplet-years of data as of 25 February 2011; other time periods had the full 5.0 triplet-years.]


Funding: The authors gratefully acknowledge the support of the United Kingdom Department of Environment, Food and Rural Affairs (Defra; http://www.defra.gov.uk/i...) for this work. CAD thanks the United Kingdom Medical Research Council (http://www.mrc.ac.uk/inde...) for Centre funding.

No competing interests declared.

RE: Analysis of further data (to 25 February 2011) on the impacts on cattle TB incidence of repeated badger culling

Clarity replied to ChristlDonnelly on 12 Apr 2011 at 07:14 GMT

Would you by any chance be able to supply all the calculated culling effect values for the adjoining lands 2 kms outside the proactive and survey-only areas?

I see analysed data now covers in excess of 5 years after culling stopped.

Have you checked that the variations in the offset for each year over 5 years prior to culling in herd incidence between lands adjoining the proactive areas and lands adjoining the survey-only areas are small in comparison to these? Obviously, what I mean by herd incidence here is an incidence weighted by herd number.

If you have not, perhaps it would be worth checking these in addition to the yearly pre-cull offsets for the proactive areas.

As far as I am aware only data in the year preceeding culling and also combined data over 3 and 10 years pre-cull have been examined and reported.

Would these analyses be effective at revealing any large annual variation in these offsets?

No competing interests declared.

RE: Analysis of further data (to 25 February 2011) on the impacts on cattle TB incidence of repeated badger culling

MattJHodgkinson replied to ChristlDonnelly on 12 Apr 2011 at 17:26 GMT

As a note to readers, the additional results reported in this comment have not been peer reviewed or otherwise assessed by PLoS ONE. They are not to be considered to be published in PLoS ONE.

Competing interests declared: I am an Associate Editor at PLoS ONE.

RE: RE: Analysis of further data (to 25 February 2011) on the impacts on cattle TB incidence of repeated badger culling

Clarity replied to MattJHodgkinson on 13 Apr 2011 at 15:29 GMT

I feel uncomfortable with the way in which I phrased my questions in the comment I made on the 12th (yesterday) for several reasons.

Perhaps I should have said that in view of the answers which you kindly gave on the 18th March, and in light of the post-cull data after month 36, sizeable 6-monthly fluctuations are not only seen in the adjoining areas (months 19-24) but also in the proactive areas (months 31-36).

Would a check on annual (or 6-monthly) variation in the pre-cull offset (i.e. historical offset), help to guage how seriously variations in these post-cull effects should be taken and perhaps establish whether or not there is a strong basis for presenting the post-cull data differently? For example, perhaps this exercise will give guidance, which would be additional to your statistical analysis, on whether it is of any value to present the data every 6 months instead of 12.

As perhaps an added bonus, a more thorough pre-cull analysis may bring to light important issues which you are not currently aware of. Perhaps for instance, it may reveal that as a result of culling badgers, sizeable errors are introduced by the way in which historical incidence is used in the analysis to calculate the presented culling effects.

No competing interests declared.