Memorandum submitted by J Thornton BSc,
BVetMed, MRCVS (BTB 18)
EXECUTIVE SUMMARY
1. I am a new graduate of the Royal Veterinary
College, and I am currently working at a mixed practice in Worcestershire.
During the last two years of our study, my colleagues and I received
extensive training and practice in critiquing scientific papers.
I have read "An Epidemiological Investigation Into Bovine
TB, Fourth Report of the ISG on Cattle TB" in the hope it
would shed some light on the epidemiology of TB in the UK cattle
herd. There are several aspects of the study that concern me,
especially since it is part of the scientific basis from which
future TB policies will be formed.
2. First, there are several flaws in data
collection (largely due to the Foot & Mouth Disease outbreak)
and there is a lot of variation in culling procedure within the
treatment groups. Professor Bourne is quick to argue that policies
must be based on science, yet many generalized and sweeping statements
are made throughout the paper that are not based on scientific
fact. Professor Bourne contradicts himself in different sections
of the paper, and certainly doesn't take criticism well from the
Independent Review Committee, whose suggestions seemed fairly
common sense (section 1.18). In fact, he seems almost biased towards
exonerating badgers instead of finding the truth about their role
in the TB disease complex. Also, the "results" that
have been given for the reactive treatment areas are presented
as solid scientific fact, when in reality the statistics are not
all that strong. As an aside, it is worrying that these same "facts"
have been used in recent RSPCA ads to influence public opinion
that badgers are innocent victims in the TB debacle. The fact
that badgers can be and are infected with bovine TB means they
inevitably play a role in the disease complex, however big or
small.
3. Overall, the design of the study is good
and well thought out. However, the study has not been implemented
and carried out well. This has allowed even more variation within
the treatment areas (which could have been better limited despite
the FMD outbreak). This variation can have large effects on the
resulting data, and no amount of statistical analysis can make
good results come from poor data. I believe that the interpretation
and use of the results of this study in creating TB culling policies
need to be done very carefully and with a large pinch of salt.
1. FMD DISRUPTED
CULLING
In section 1.5, page 9, Bourne states that he
does not believe the FMD epidemic had a large impact on the trial's
integrity and outcome. I disagree. The aim of the study was to
determine "the effect on the incidence of bovine TB of two
different approaches to badger culling, each of which represented
a potential practical policy option" (section 1.4, page 9).
The very fact that FMD disrupted and delayed culling in the proactive
and reactive culling areas completely debases the foundation of
the study by allowing a source of variation to come into the treatment
groups.
1.1 Variation in Proactive Area Culling
In the proactive areas, badger densities were
to be reduced to low levels and maintain this low density by culling
every five to nine months. Seven of the 10 proactive areas received
the intital cull and were active before FMD. However:
(a) 5/10 proactive areas had >1.5 years
delay between intial and first culling; three of these five were
>2 years (see table 3.1, page 24); and
(b) only two of the 10 areas had recommended
five to nine month wait between initial and first cull as stated
in standard operating procedures in section 3.16, page 23.
Bourne argues that the seven areas activated
before FMD continued to "accumulate functional data"
during FMD "because badger numbers had been reduced".
Later on in sections 3.19-3.21, several of the proactive areas'
first follow up culls resulted in more badgers being culled than
in the intial cull. This is explained by the seasonality of the
culls, but is it possible for the badger population to have increased
in the 25 month delay between the intial and first follow-up culls?
1.2 Variation in Reactive Area Culling
In the reactive areas, badgers were supposed
to be culled when there was a notification of a TB herd breakdown.
However:
(a) 36% of notifications were not culled,
mostly due to some ministerial announcement (section 3.25, page
27). That is over a third of the notifications received in the
reactive areas!
(b.) Of the notifications that were culled,
some had full and others partial culls.
(c) No culling of badgers was allowed during
the "closed season", so if a notification was received
during these months, culling was delayed.
(d) The numbers of badgers culled in the
reactive areas were very small compared to the numbers culled
in the proactive areas (Appendix D, Table 3.5 vs Figure 3.4).
Does culling 100 badgers over three years in triplet A really
going to make any difference?
(e) Bourne admits in Section 3.61, page 38
that "delays to reactive culling imposed by the closed season
have been linked to the failure of this experimental treatment
to reduce the incidence of cattle TB."
1.2.1 This is far too much variation within
the reactive treatment groups. The resulting data from the reactive
areas are then lumped together for statistical analysis. Is it
any wonder with no culling, partial culling, and full culling
in this treatment group resulted in such wide and imprecise confidence
intervals (section 3.41, page 33)? I can't see how any useful
information can be obtained from the reactive areas in this trial.
The whole point of this treatment group was to determine the effect
of reactive culling on incidence of TB in cattle herds, and every
notification should have been culled as similarly as possible.
1.3 Seasonality affected numbers culled
It is obvious that season has an effect on badger
activity and thus the number of badgers caught for culling. This
is another source of variation within and between the proactive
and reactive treatments.
1.4 Badger densities difficult to measure
It was difficult to measure the effectiveness
of badger culling carried out in the course of the RBCT (section
3.30; page 30).
— There is no precise way to measure
badger density.
— Quantification of badger field signs
are imprecise.
2. GENERALIZED
AND SWEEPING
STATEMENTS
2.1 "Clear Conclusions" From Reactive
Trial Areas
Bourne declares in section 3.42, page. 33 that
"The clear conclusion supported by all the analyses undertaken
by the ISG is that there is convincing evidence that reactive
culling of badgers does not offer a beneficial effect large enough
to make it useful as a practical policy option and that there
is substantial but not overwhelming evidence of an adverse effect
of the reactive strategy."
(a) The evidence is far from convincing and
the conclusions far from clear!
(b) Please refer to paragraph 1.2 of this
letter. Any results from the reactive treatment areas will be
as varied as the culling within this treatment group. Therefore,
how can reactive culling (or should I say no culling in 1/3 of
notifications, partial culling or full culling) be associtated
or linked with anything, let alone an increase incidence of TB
in cattle?
(c) The wide confidence intervals accompanying
the supposed increase in TB incidence in reactive areas verifies
the imprecise and varied results.
(d) Who determines whether "a beneficial
effect" is or is not "large enough"? Statiticians
playing with data? Pro-badger groups? Farmers shut down due to
TB in their herds? Policy-makers pressured by all of the above?
(e) It does not matter how "extensive
and carefully conducted" statistical analyses are carried
out. If you put garbage data in you will get garbage results out.
This is why the way a study is designed and performed is so very
important!
2.2 Ignoring the Wildlife Component of TB
I am concerned about the statement made by Professor
Bourne in section 9.9, page.62, that suggests focussing only on
cattle factors and ignoring the wildlife components could lead
to improvement of TB control.
(a) This is an unsupported and dangerous
suggestion, especially from someone who is so concerned that TB
policies should be based on good science (section 1.18).
(b) TB disease complex is influenced by many
different factors, and wildlife acting as TB resevoirs are involved
in this complex.
(c) Badgers are without doubt acting as wildlife
resvoirs for the disease, simply because it is known they can
be and are infected with bovine TB. Bourne states this as well
(section 1.18).
(d) In Appendix D, data on the badger removal
operations in trial areas under the interim strategy show that
a high percentage of the badgers caught were infected with TB.
This ranged from 13-76% (mean = 26%, median = 23%).
2.2 1 That is a quarter of the badgers removed
infected with TB!! If any disease had such a high incidence in
the human population, there would be mass hysteria!
2.2 2 How can Bourne suggest that the wildlife
component of this disease complex be ignored?! Even if cattle
to cattle transmission is found to be the main way the disease
is spread, surely such a high percentage of TB in the badger population
should be addressed, for badger welfare if nothing else.
3. THE RBCT IS
NOT GOOD
SCIENCE
3.1 Policy Should Be Based on Good Science
Professor Bourne is insistent that good science
form the basis of TB policy-making. In Section 1.18, page 15,
Bourne says "it is obvious that badgers are `implicated'
simply because we know that they are susceptible to bovine TB
infection and therefore inevitably a part of the overall disease
complex; but that does not suggest what, if anything, should be
done about it." I agree with Professor Bourne on this point,
however I don't believe the RBCT has shed any light on badgers
contribution in the disease complex, and certainly the "benefits
of the trial's findings" will not aid the Ministers in forming
policy decisions.
3.2 Hidden and Exposed Results
He was agast that the Independent Review Group
recommended "premature release of trial data, an action that
would violate the generally accepted principles of conducting
trials of this sort while offering no clear guidance as to how
policy might be improved." Yet in section 3.39, in regards
to the apparent increased incidence of TB in reactive culling
areas, Bourne and colleauges were "obliged to bring this
information to the attention of Ministers". Results from
the proactive treatment group were not reported on.
3.3 Public Manipulation
Why is it that the data from the proactive treatment
remains strictly confidential at this time when the reactive data
"results" have been freely exposed? Why is it that these
same results and conclusions have been used in RSPCA ads to sway
public opinion? Not only are the conclusions from the reactive
treatment based on a poorly conducted trial, they are now being
used to incite public opinion against farmers and for badgers
in the hopes that the public will pressure policy makers not to
cull badgers. These tactics are appalling! This is not science
and wisdom guiding policy. This is manipulation.
4. LET'S
LEARN FROM
THE PAST
4.1 Past UK TB Epidemic
It is important to have some common sense and
look to the past for guidance in making TB policies. According
to the DEFRA website, the outbreak of cattle TB that occured in
the 1950's was controlled by a compulsory test and slaughter policy
(using the tuberculin intradermal skin test that Professor Bourne
has rubbished in his Fourth Report). By 1980 the incidence of
TB in the UK herd was low.
4.2 What has changed?
So what has changed that this same method is
not working to control TB today?
(a) Badgers became a protected species in
1973, and further regulations in 1975 and 1980 increased this
protection.
(b) The UK badger population today is estimated
between 250,000-300,000. Badgers like using well-trodden paths;
they like eating maize and other food sources readily available
on farms; they use straw to line their burrows; they prefer to
live near wooded areas on farmland;
(c) The badger population is infected with
bovine TB.
(d) I feel that policy needs to carefully
weigh the implied impact of the badgers' role as a wildlife TB
resevoir.
4.3 Other countries
As far as I am aware, only Great Britain and
Ireland have a rampant TB problem in their national herds today.
(a) What have other countries done to control
TB in their cattle herds?
(b) Was there a wildlife resevoir contributing
to the problem?
(c) Was a culling policy initiated and did
it make any difference?
February 2006
| 
